Full text of "Vogels Textbook Of Quantitative Inorganic Analysis"

See other formats

Vogel’s

Textbook of Quantitative
Inorganic Analysis

The English Language Book Society is funded
by the Overseas Development Administration
of the British Government. It makes available
low-priced, unabridged editions of British
publishers’ textbooks to students in developing
countries. Below is a list of some other books
on chcmisliy published under the ELBS
imprint.

Atkins

Physical Gicmistry
Oxford University Press

Burgess

Metal Ions in Solutions
Ellis Horwood

Finar

Organic Chemistry Vols 1-2
Longman

Furniss eroA (revisers)

Vogel’s Textbook of Practical Organic

Chemistry

Longman

Lee

A New Concise Inorganic Chemistry
Van Nostrand Rcinhold

Li pi rot

Modem Inorganic Chemistry
Bell & Hyman

Liptrot, Thompson and Walker
Modem Physical Chemistry
Bell & Hyman

McWeeny
Coulson’s Valence
Oxford University Press

Sharpe

Inorganic Chemistry
Longman

Wells

Structural Inorganic Chemistry.

Oxford University Press

Vogel’s

Textbook of Quantitative
Inorganic Analysis

Including Elementary Instrumental Analysis
Fourth Edition

Revised by

J. BASSETT MScCChemFRIC

Senior Lecturer in Inorganic Chemistry, Thames Polytechnic

R. C. DENNEY BSc PKD CChem FRIC
Senior Lecturer in Inorganic Chemistry, Thames Polytechnic

G. H. JEFFERY BSc PhD CChem FRIC

Formerly Principal Lecturer and Deputy Head of the School of Chemistry,
Thames Polytechnic

J. MENDHAM MScCChemMRIC

Senior Lecturer in Analytical Chemistry, Thames Polytechnic

English Language Book Society/Longman

Longman Scientific & Tcchnica!

Longman Group UK Ltd

Longman House. Burnt Mill, Harlow,

Essex CM20 2JE. England

All rights reserved; no part of this publication
may be reproduced, stored in a retrieval
system, or transmitted in any form or by any
means, electronic, mechanical, phottKopying,
recording, or otherwise, without the prior
written permission of the Publishers.

First published 1929
Second edition 1951

Third edition (publi.shcd under the title
A Textbook of Quanlitalirc Inorgaiuc Analysts
Including Elementary Instrumental Analysts)
1961

Fourth edition 1978

Reprinted 1979, 1981, 1983, 1985, 1986

ELBS edition first published 1962
Reprinted 1964, 1968, 1969. 1971, 1974, 1975
ELBS edition of fourth edition 1978
Reprinted 1979, 1982, 1985, 1986

ISBN 0 582 44663 5

Printed in Great Britain by
William Clowes Limited
Beccles and London

CONTENTS

PART A FUNDAMENTALS OF QUANTITATIVE INORGANIC ANALYSIS
Chapter I Introduction 3

. a

I, 1 Chemical analysis

1, 2 Sampling ^

I, 3 Types of analysis ^

1.4 Use of literature

1. 5 Common techniques ^

1 . 6 Other techniques ‘6

1. 7 Instrumental methods 7

1.8 Time and money, accuracy and range 7

1. 9 Interferences 8

1. 10 Acc uracy and precisio n . .10

1.11 Summary^ 10

1.12 Selected bibliography 11

Chapter n Fundamental theoretical principles 14

11.1 Electrolytic dissociation 14

11.2 The law of mass action 17

11.3 Activity and activity coefficient ' 18

11.4 Acid-base equilibria in water 19

II, 5 Strengths of acids and bases 20

II, 6 Dissociation of polyprotic (polybasic) acids 22

11,7 Common ion effect .23

II, 8 Solubility product ' 25

II, 9 Quantitative effects of a common ion . 27

11,10 Fractional precipitation 29

II, 1 1 Complex ions 30

II, 12 Effect of acids upon the solubility of a precipitate 32

11,13 Effect of temperature upon the solubility of a precipitate 33

II, 1 4 Effect of the solvent' upon the solubility of a precipitate , 33

11,15 The ionic product of water 33

II, 16
II, 17
II, 18
II, 19
II, 20
11,21
II, 22
II, 23
II, 24

11.25

11.26

The hydrogen-ion exponent, pH
The hydrolysis of salts

Hydrolysis constant and degree of hydroK^is
Buller solutions
Electrode potentials
Concentration cells

Calculation of thee.m.f. of a voltaic cell

vJxidalion-rcduction cells

Calculation of the standard reduction potential

Chapter HI Common

1 Introduction

apparatus and basic techniques

53
56

Balances

JJU The analytical balance

III 4 of the analyt.cal bala

nr reference masses

III 7 balance

.7 Top loading balances

rtt’o ”*°‘^“'onic balances

I no

errors in weighing

Graduated glassivarc
II. 12 Units of volume

III 4

III 5 ?"^“^“^'i«PP<->ratus

,5 Graduated nasfcs

JJI, 16 Pipettes
III- 17 Burettes

m’

m’^n tiurcttes

• 20 Graduated (measuring) cylinders

W-aerforjahoratorynse

II ’yi water

22 Wash bottles

General apparatus

m 25 P'^Pp^rSs'""'’

• ‘ boxes

60
60
61
62

68
69

74
76
78
81
SI
82

85
88
91

111,27 Stirring apparatus
III, 28 Filtration apparatus
III, 29 Weighing bottles

Reagents and standard solutions

III, 30 Reagents

III, 3 1 Purification of substances

III, 32 Preparation and storage of standard solutions .

99
100

Some basic techniques

III, 33 Preparation of the substance for analysis

III, 34 Weighing the sample

III, 35 Solution of the sample

III, 36 Precipitation

III, 37 Filtration

III, 38 Filter papers

III, 39 Filter pulp

III, 40 Gooch crucibles

III, 4 1 Crucibles with permanent porous plates

III, 42 Washing of precipitates

111,43 Technique of filtration

III, 44 Drying and ignition of precipitates

111,45 Perforated screens for crucibles

III, 46 The Schdniger oxygen flask method for elemental analysis’
III, 47 References . ‘

III, 48 Selected bibliography

103

104
104
106
107
107
109
109

,111

,112

112

115

116
117

PARTB ERRORS AND SAMPLING

Chapter IV Errors and statistics 121

IV, 1 Limitations of analytical methods 121

IV, 2 Accuracy 121

IV, 3 Precision • 122

IV, 4 Classification of errors ■' 123

IV, 5 Minimisation of errors ‘124

IV, 6 Significant figures and computations 125

IV, 7 Mean (average) deviation. Standard deviation ' 127

IV, 8 Normal (Gaussian) distribution ’ ■ . • 129

IV, 9 Comparison of results . • , , j 2 q

IV, 10 The number of parallel determinations' ' ’ ‘ 132

IV, 11 The value of statistics ' ' 133

IV, 12 References ' ' - 233

IV, 13 Selected bibliography ,, 234

Chapter V Sampling 235

V, 1 The basis of sampling : , 235

vii

V, 2 Sampling and physical slate

V, 3 Crushing and grinding

V, 4 Hazards in sampling

V, 5 References

V. 6 Selected bibliography

135

137

138
138
138

PARTC SEPARATIVE TECHNIQUES

Introduction

Chapter VI Solvent extraction

1 Genera! discussion

VI. 2 Factors favouring solvent extraction

VI, 3 Quantitative treatment of solvent extraction equilibria
VI. 4 Ion-association complexes

VI, 5 Extraction reagents

VI, 6 Some practical considerations

Some applications

VI, 7 Determination of beryllium as the acety lacetonc complex
Determination of boron using ferroin
Detennination of copper as the diethyldithiocarbamate
complex

Determination of copper as the 'nco-cuproin' complex
Determination of iron by chloride extraction
Determination of iron as the 8-hydroxyquinoIaie
Determination of lead by thedithizone method
Determination of molybdenum by the thiocyanate method
Determination of nickel as the dimcthylglyoximc complex
Determination of silver by extraction as its ion-association
complex with 1.10-phenanthrolinc and bromopyrogallol red
Detennination of uranium as the S-hydroxyqumoIatc
References

Selected bibliography

VII Ion exchange

^^11,1 General discussion

VII, 2 Action of ion exchange resins
Ion exchange chromatography
Ion exchange in organic and aqueous-organic solvents
Chelating ion e.xchange resins
Liquid ion exchangers

VI, 8
VI, 9

VI, 10
VI, 11
VI. 12
VI, 13
VI, 14
VI, 15
VI, 16

VI, 17
VI, 18
VI, 19

VII, 3
VII, 4
VII, 5
VII, 6

Applications in analytical chemistry
VII, 7 Experimenialtcchniques

VII, 8 Determination of the capacity of an ion exchange resin
(column method)

viii

14!

143

145

146

148

149
151

153

154

155

156
1.57
158
158
160
161

162

163

165

167

172

175

177

178
178

180

VII, 9 Separation of zinc and magnesium on an anion exchanger
VII, 10 Separation of chloride and bromide on an anion exchanger
VII, 1 1 Determination of the total cation concentration in water
VII, 1 2 Separation of cobalt and nickel on an anion exchanger

VII, 1 3 Separation of cadmium and zinc on an anion exchanger
VII, 14 Determination of fluoride with the aid of a cation
exchanger

VII, 1 5 Determination of sulphur in iron pyrites with the aid of a
cation exchanger

VII, 1 6 Separation of cobalt and uranium from mixed

aqueous-organic solvent using a cation exchange resin
VII, 17 Determination of uranium with the aid of a liquid anion
exchanger

VII, 18 Concentration of copper(II) ions from a brine solution
using a chelating ion exchange resin
VII, 19 References
VII, 20 Selected bibliography

■ 181
182
183
183

185

186

187

188

189

190

191
191

Chapter Vin Paper, thin layer and column
chromatography

VIII, I General introduction,

VIII, 2 Thin Layer Chromatography

VIII, 3 High Performance Liquid Chromatography

VIII, 4 Separation of nickel, manganese, cobalt and zinc and

determination of Rp values

VIII, 5 Separation of nickel, copper, cobalt, and zinc

VIII, 6 Semi-quantitative separation of copper, cobalt, and nickel

on slotted paper strips

VIII, 7 Separation of iron and aluminium on a cellulose column

VIII, 8 Separation of cobalt and nickel on a cellulose column

VIII, 9 Separation of copper and nickel on a cellulose column

VIII, 10 References
VIII, 1 1 Selected bibliography

apter IX Gg^hromatography

IX, 1

Introduction ^

IX, 2

Apparatus

IX, 3

Programmed-temperature gas chromatography

IX, 4

Quantitative analysis by GLC

IX, 5

Gas chromatography of metal chelates

IX, 6

Determination of aluminium by gas chromatographic

analysis of its tris(acetylacetonato) complex

IX, 7

References

IX, 8

Selected bibliography

193

196

198

199

200

201

202

205

206
207
207

209

214

216

218

219

PART D TITRIMETRY AND GRAVIMETRY

Chanter X Titrimclric analysis 223

2'’3

A •nicorctical considerations

X, 1 Tilrimetric analysis .

X, 2 Classification of reactions in tiifimctric analysis

X. 3 Standard solutions 225

X, 4 Advantages of the use of the equivalent system 234

X, 5 Preparation of standard solutions 235

X, 6 Primary standard substances 235

A.l Theory of acid-base titrations 236

X, 7 Neutralisation indicators 236

X, 8 Preparation of indicator solutions 242

X, 9 Mixed indicators 243

X, 10 Universal or rnutliplc range indicators 244

X, 1 1 Neutralisation curv'cs 244

X, 1 2 Neutralisation of a strong acid and a strong ba«;e 244

X, 13 Neutralisation of a sveak acid with a strong base 247

X, 14 Neutralisation of a svciik base vs'ith a stiong acid 249

X, 1 5 Neutralisation of a weak acid with a weak base 251

X, 16 Neutralisation of a polyproticactd with a strong base 251

X, 17 Titration of anions of weak adds (Hronsted bases) with

strong acids. ‘Displacement titrations’ 253

X, 18 Choice ofindicalor.s in neutralisation reactions 255

A.l Theory of compk.vaiion titrations 257

X, 19 Introduction 257

X, 20 Stability of complexes 2.57

X, 2 I Factors influencing the stability of compic.ve.s 258

X, 22 A simple complcxation titration 260

X, 23 Complexoncs 261

X, 24 Stability constants of FDTA complexes 264

X. 25 Titration curves 265

X, 26 Types of EDTA titrations 266

X, 27 Titration of niixtiirc.s. selectivity, masking and denvasking

agents 267

X, 28 Mctalion indicators 269

A.3 Theory of precipitation titrations 279

X, 29 Precipitation reactions 279

X, 30 Determination of end-points in precipitation reactions 28 1

A.4 Theory- of oxidation-reduction titrations 288

X, 31 Change of the electrode potential during the titration of a

reductanl with an oxidant 288

X, 32 Formal potentials 291

X, 33 Detection of the end point in oxidation-reduction
titrations '

B Experimental details 296

B.l Aqueous acid-base titrations (Acidimetry and alkalimetry) 296

X, 34 Preparation of a standard acid 296

X, 35 Preparation of constant-boiling-point hydrochloric acid 297

X, 36 Direct preparation of 0. 1 M-hydrochloric acid from the

constant-boiling-point acid • ' 297

X, 37 Preparationof approximately 0.1 M-hydrochl6ric acid '

and standardisation - , 298

X, 38 Preparation of standard alkali 301

X, 39 Standardisation of the approximately 0. 1 Af-sodium

hydroxide ^ 304

X, 40 Other standard substances for acidimetry and alkalimetry 305

X, 41 Standard barium hydroxide (baryta) solution 306

X, 42 Determination of the NajCOj content of washing soda 306

X, 43 Determination of the strength of concentrated acids 308

X, 44 Determination of a mixture of carbonate and hydroxide.

(Analysis of commercial caustic soda) 309

X, 45 Determination of a mixture of carbonate and hydrogen-

carbonate 310

X, 46 Determination of boric acid 311

X, 47 Determination of ammonia in an ammonium salt 312

X, 48 Determination of nitrates 314

X, 49 Determination of phosphate (precipitation as quinoline

raolybdophosphate) 314

B.2 Complexation titrations 316

X, 50 Standard EDTA solutions , . 317

X, 51 Some practical considerations 317

Determination of cations 319

X, 52 Determination of aluminium ; back titration using

Solochrome Black indicator , ' , 319

X, 53 Determination of barium : direct titration with Methyl

Thymol Blue indicator 319

X, 54 Determination of bismuth ; direct titration using Xylenol

Orange indicator , , ... 320

X, 55 Determination ofcalcium: substitution titration using .

Solochrome Black (Eriochrome Black T) indicator . 320

X, 56 Determination ofcopper: direct titration using Fast ^

Sulphon Black F indicator ■ ,321

X, 57 Determination of iron(III) ; direct titration using .

Variamine Blue indicator ; 322

X, 58 Determination ofnickel: direct titrations using (a)

Murexide and (6) Bromopyrogallol Red as indicators • 322

X, 59 Delerminaiion of silver; indireci .. •

...

Analysis of mixtures of cations

;

X, 63 Determination ofctilcinm m

using EGTA as titrant P^'-’^unce of magnesium

X,64 Determination ofthcioi-it I, I

tcmporanOofwiterii -’ ' c *i "‘■’■‘'''^P‘-'™-‘'>unt and

Bbck T)?„dic,,;; ' llhct

“ pSZ! n° ;r “'"E

admtal'reu™ of n”or”“'“"‘' "ml nnc i„

“ RMaminalion ofcl„omiom("i',V‘

® Celoraooaiion of

«rferoma„p,„o, »l>o„ : o„ol,,is

Determination of nicfel in

nickelsteel Ton: analysis of

admixenre: Tn" S'„r and lood

;;SZ;:r“"'”'"“''»w.n,n,,,„,„do,o„d

crniination of sulp|,a(e.s
J.'” sSSto„“'T"l«-d,ra,c

X,’ 79

method by an indirect

323

324

325

327

328

329

330

331

332

333

334

335

336

337
33S
339

339

340

X, 82 Preparation and use of 0. 1 Af-ammonium or potassium

thiocyanate. Titrations according to Volhard’s method * ' 340

X, 83 Determination of silver in a silver alloy '342

X, 84 Determination of chlorides (Volhard’s method) 342

X, 85 Determination of fluoride ; precipitation as lead

chlorofluoride coupled with Volhard titration 343

X, 86 Determination of arsenates • 344

X, 87 Determination of cyanides 345

X, 88 Determination ofchlorides by titration with mercury(II)

nitrate solution 346

X, 89 Determination of potassium 347

B.4 Oxidation— reduction titrations 348

Oxidations with potassium permanganate 348

X, 90 Discussion 348

X, 91 Preparation of 0. 1 V-potassium permanganate 351

X, 92 Standardisation of permanganate solutions ■ 351

X, 93 Determination of iron(II) 354

X, 94 Determination of calcium • ■ 354

X, 95 Analysis of hydrogen peroxide 355

X, 96 Determination of manganese dioxide in pyrolusite .> 356

X, 97 Determination of nitrites ' " 356

X, 98 Determination of persulphates .■ 357

X, 99 Determination of manganese in steel 358

Oxidations with potassium dichromate 359

X, 100 Discussion ■ 359

X, 101 Preparation of 0.1 A potassium dichromate . 360

X, 102 Standardisation of potassium dichromate solution against

iron 360

X, 103 Determination of iron(II) ■ 360

X, 104 Determination of chromium in a chromium(III) salt 361

X, 105 Determination of chromium in chromite ' 362

X, 106 Determination of chlorate 363

Oxidations with cerium( IV) sulphate solutions 363

X, 107 General discussion ' 363

X, 108 Preparation of O.lVcerium(IV) sulphate , 365

X, 109 Standardisation of cerium(IV) sulphate solutions 366

X, 110 Determination of copper , . , 367

X, 1 1 1 Determination of molybdate 367

X, 112 Determination of tellurite 368

X, 113 Determination of cerium(III), 368

X, 114 Determination of nitrites ' ! ' 369

Oxidation and reduction processes involving iodine. lodometric titrations 370

X, 1 1 5 General discussion , ^70

xiii

X, 116
X.I17
X, 1)8
XJI9
X, 120
X, 121
X, 122
X. 123
X, 124
X, 125

Deleclion oftlicenci-poini
Preparation ofO.lA'-sodiiini tliiosulphaie

i reparation oro.JA'-iodincsohiiion

Standardisation of iodine solutions

Determination of copper in anTrc
'^'"'termination ofcliloratcs
AnalysLs of hydropen peroxide

lJcicnmnationoftiieaviil-.M,.,-M ■

powder ‘‘'•"Dble chlorine m bleaching

Determination ofarscnic(V)

General discussion
Preparation of 0 0'’S ir

of „v;

P«orn,|naiionori,j.,t,„|„5
D«=niiinai,onofvj„a*„5,

Kli,

X, 139
X, 140

X, 126
X, 127
X, 128
X, 129
X, 130

X, 132
X, 133
XJ34
X, 135
X,136

Preparation ofO .
Determination of ani^ ' •’’■f’oiate
polcrminalioa ofTO„7°J!v

X, 141

r/ j '.^^roxyian

iFs"'-?--

X, 144
X,145
X, 146
X,147
X, 148

— -..icussion

Se'<=cled bibliography

Chapter XI Gnvim
XU ra. . "‘"'"nefry
XJ,2
XI, 3
XI, 4

xiv

J‘eco»„MaU„„

^upersaturation

''"‘''’r«T/.a,or„™,a„„„

372

374

375

377

378

379

380

381
38!

382

383

384

385

386

387

388

389

390

392

393

394

395

397

398

399

401

402

403

404
407

XI, 5

The purity of the precipitate. Coprecipitation

409

XI, 6

Conditions of precipitation

410

XI, 7

Precipitation from homogeneous solution

411

XI, 8

Washing of the precipitate

413

XI, 9

Ignition of the precipitate. Thermogravimetric method of • .

analysis -

. 414

Quantitative separations based upon precipitation methods

415

XI, 10

Fractional precipitation

415

XI, 11

Organic precipitants

419

XI, 12

Volatilisation or evolution methods

430

Practical gravimetric analysis

431

XI, 13

General discussion

431

XI, 14

Calculations of gravimetric analysis

431

Simple gravimetric determinations

432

XI, 15

Determination of water of hydration in crystallised barium

chloride

432

XI, 16

Other determinations by ignition

' 432

XI, 17

Determination of chloride as silver chloride

433

XI, 18

Determination of aluminium as aluminium oxide

435

XI, 19

Determination of aluminium as the 8-hydroxyquinolate,

Al(C 9 HgON) 3 , with precipitation from homogeneous

solution

436

XI, 20

Determination of calcium as oxalate

437

XI, 21

Determination of iron as iron(III) oxide

440

XI, 22

Determination of lead as chromate

444

XI, 23

Determination of magnesium as the ammonium phosphate

hexahydrate and as the pyrophosphate

444

XI, 24

Determination of nickel as the dimethylglyoximate

' 447

Systematic gravimetric analysis

449

XI, 25

General discussion

, 449

Cations

XI, 26

Aluminium

XI, 27

Ammonium

XI, 28

Antimony

XI, 29

Arsenic

XI, 30

Barium

XI, 31

Beryllium

XI, 32

Bismuth

XI, 33

Cadmium

XI, 34

Calcium

XI, 35

Cerium

XI, 36

Chromium

XI, 37

Cobalt

449

449
,450

450

451

452
.454

455

456
458

458

459

460

•XV

XI, 38 Copper
XI, 39 Gold
XI, 40 Iron
XI, 41 Lead
X!,42 Lithium
XL 43 Magnesium
XL 44 Manganese
XL 45 Mcrcur>'

XL 46 Molybdenum

XL 47 Nickel

XL 48 Palladium

XL 49 Platinum

XL 50 Potassium

XL 51 Selenium and tellurium

XL 52 Silver

XL 53 Sodium

XL 54 Strontium

XL 55 Thallium

Xl, 56 Thorium

XL 57 Tin

XL 58 Titanium

XL 59 Tungsten

Xl, 60 Uranium

XL 61 Vanadium

XL 62 Zinc

XL 63 Zirconium

462

464

465

467

468

469

470

471

473

474

475
477
479
479
481
4S2

483

484

485

486

487

488
488
490

Anions

491

XL 64

Borate

491

XL 65

Bromatc and bromide

491

XL66

Carbonate

492

XL 67

Chlorate

493

XL 68

Chloride

494

XL 69

Cyanide

494

XL 70

Fluoride

494

XL71

Fluorosiiicale

495

XL 72

Hcxacyanofcrratc( ill)

496

XL 73

Hcxacyanoferraiclll)

496

XL 74

Hypophosphitc

496

XL 75

lodate

496

XL 76

Iodide

496

XL 77

Nitrate

497

XL 78

Nitrite

497

XL 79

Oxalate

498

XL 80

Perchlorate

498

XL 81

Phosphate

498

XVI

XI, 82 Phosphite

XI, 83 Silicate ■ 501

XI, 84 Sulphate 504

XI, 85 Sulphide 507

XI, 86 Sulphite 510

XI, 87 Thiocyanate 510

XI, 88 Thiosulphate ' ■ 510

XI, 89 References 511

XI, 90 Selected bibliography 511

PARTE ELECTROANALYTICAL METHODS

Chapter Xn Electro-gravimetry 515

XII, 1 Theory of electro-gravimetric analysis 515

XII, 2 Electrode reactions 518

XII, 3 Overpotential 519

XII, 4 Completeness of deposition 521

XII, 5 Electrolytic separation of metals 522

XII, 6 Character of the deposit 523

XII, 7 Electrolytic separation of metals with controlled cathode

potential .524

Electrolytic determinations at constant current 521

XII, 8 Apparatus 527

XII, 9 Copper 532

XII, 10 Lead 535

XII, 1 1 Cadmium 536

XII, 12 Silver -537

XII, 1 3 Electrolytic separation and determination of copper and

nickel 537

Electrolytic determinations with controlled cathode potential 538

XII, 14 Antimony, copper, lead, and tin in an alloy (e.g., bearing

metal) 539

XII, 15 Internal electrolysis 540

XII, 16 References 54 O

XII, 17 Selected bibliography 541

Chapter Xin Coulometry 542

XIII, 1 General discussion . ' 542

XIII, 2 Coulometry at controlled potential , 543

XIII, 3 Separation of nickel and cobalt by coulometric.analysis at .

, controlled potential ,

Coulometry at constant current : coulometric titrations • 548

XIII, 4 General discussion .

xvii

XIII, 5
XIII, 6
XIII, 7
XIII, 8

XIII, 9
XIII, 10
XIII, 1 1
XIII, 12
XIII, 13
XIII, 14
XIII, 15
XIII. 16
XIII. 17
XIII. 18
XIII, 19

Instrumentation

Antimony(III)

Tiiiosulpliate

Oxine (S-hydroxyquinoiinc)

klichromatc ion)

Chloride, bromide, and iodide
Bromide and iodide
Titration of acids
Titration ofbases
References

Selected bibliography

Chapter XIV Potenfiometry
^TV, 1 Introduction

Reference electrodes

XIV 3 '>>'^‘rogen elect rode

XV4 If'=«'°nicl electrode

. ‘'"^'’''cr-silver chloride electrode

Micator electrodes
XIV discussion

Xiv’ 7 I’ydrogcn electrode

Xiv's electrode

The glass electrode
^^^ensitive electrodes

''i UquTdr

^''^’>'>mtentationand, ‘’'e electrodes

XIv’k P^^'^‘-‘lcrs

’ ' ion meters

^•!^‘^<Poientio,ne,ry

''‘^'‘^'^■"••^‘■•onof/luoridc

551

552
555

556

557

558

559

560
560
562
562
564

564

565

566

568

569
571

571
57]

572

573

575

577

578

579
579
582
584

585

589

591

XIV, 19 Use ofbimetallic electrode systems 592

XIV, 20 Polarised indicator electrodes 593

XIV, 21 Differential potentiometric titration 593

XIV, 22 Automatic potentiometric titrations 594

XIV, 23 Location of end points 596

XIV, 24 Some general considerations 600

XIV, 25 Some experimental details for potentiometric titrations 602

XIV, 26 Determination of copper > 605

XIV, 27 Determination of chromium 606

XIV, 28 Determination of manganese 606

XIV, 29 Potentiometric EDTA titrations with the mercury electrode 607
XIV, 30 Determination of iron(III) with EDTA 610

XIV, 3 1 Standardisation of potassium permanganate solution with

potassium iodide 610

XIV, 32 Determination of nickel and of cobalt by complexation ,

with cyanide 611

XIV, 33 Determination of fluoride by a null-point method 612

XIV, 34 References 613

XIV, 35 Selected bibliography 614

Chapter XV Conductometric titrations 615

XV, 1 General considerations 615

XV, 2 The measurement of conductivity 6 1 6

Conductometric ( low frequency ) titrations 617

XV, 3 The basis of conductometric titrations 617

XV, 4 Apparatus and measurements 619

XV, 5 Applications of conductometric titrations 621

XV, 6 Some experimental details for conductometric titrations 625

High frequency titrations 626

XV, 7 General considerations 626

XV, 8 Apparatus , 628

XV, 9 Advantages of the technique 629

XV, 10 Some examples of high frequency titrations 629

XV, 1 1 References 630

XV, 12 Selected bibliography 631

Chapter XVI Voltammetry

XVI, 1 Introduction

Polarography

XVI, 2 Basic principles

Direct current polarography
XVI, 3 Theoretical principles
XVI, 4 Quantitative technique

632

633
633

636

642

XIX

XVI, 5 Evaluation of quantitative results

XVI, 6 Measurement of wave heights

XVI, 7 Manual non-recording polarographs

XVI. 8 Commercial polarographs
XVI, 9 Ancillary' equipment for polarography
XVI, 10 Determination of the halfwave potential of the cadmium
ion in Af-polassium chloride solution
XVI. 1 1 Determination of cadmium in solution
XVI, 12 Investigation of the influence of dissolved o.xygcn
XVI, 1 3 Determination of lead and copper in steel

Aliernating current polarography
XVI, 14 The nature of a.c. methods
XVI, 15 Simple a.c. polarography
XVI. 16 Square-wave polarography
XST, 17 Pulse polarography

Oscillographic polarography

XVI, 18 Controlled potential methods

XVI, 19 Controlled current methods

XVI, 20 Instrumentation

XVI, 21 Quantitative determinations

644

645

647

648
650

653

655

656

657

658
658

660

66 !

662

663

Anodic stripping voliarnniciry 664

XVI, 22 Basic principles 664

XVI, 23 Some fundamental features 665

XVI, 24 Instrumentation 667

Chronopotentiomciry 668

XVI, 25 Basic principles 668

XVI, 26 Experimental procedure 669

XVI. 27 References 670

XVI, 28 Selected bibliography 670

Chapter XVII Amperometry 672

XVII, 1 Ampcrometric titrations 672

XVH, 2 Technique of ampcrometric titrations with the dropping

mercury electrode 675

XVII, 3 Determination of lead with standard potassium

dichromate solution 676

XVll. 4 Determination of sulphate with standard lead nitrate

solution 677

XVII, 5 Determination of nickel with dinicthylglyo.ximc 678

X 11, 6 Determination of fluoride with standard thorium nitrate
solution

XVII, 7 Determination of zinc with EDTA 6S0

n, 8 Titration of an iodide solution with mcrcur\’(ll) nitrate

solution ■ /-cn

XVII, 9 Determination of potassium with sodium

tetraphenylborate (graphite indicating electrode) 681

Titrations with the rotating platinum micro-electrode 682

XVII, 10 Discussion and apparatus , , 682

XVII, 1 1 Determination of thiosulphate with iodine 683

XVII, 12 Determination of arsenite with standard iodine solution 684

XVII, 1 3 Determination of antimony with standard potassium

bromate solution 684

Biamperometric titrations 685

XVII, 14 General discussion . 685

XVII, 15 Titration of thiosulphate with iodine (‘dead-stop end-

point’) 685

XVII, 16 Determination of nitrate , 686

XVII, 17 Determination of water with Karl Fischer reagent 687

Xyil, 1 8 Determination of the water content of a salt hydrate 688

XVII, 19 Selected bibliography 690

PARTF SPECTROANALYTI CAL METHODS

Chapter XVin Colorimetry and spectrophotometry

XVIII, 1 General discussion

XVIII, 2 Theory of spectrophotometry and colorimetry
XVIII, 3 Classification of methods of ‘colour’ measurement or

comparison

XVIII, 4 Standard series method

XVIII, 5 Duplication method

XVIII, 6 Balancing method

XVIII, 7 Photoelectric photometer method

Instruments

XVIII, 8 Photoelectric colorimeters (absorptiometers)

XVIII, 9 Photoelectric spectrophotometers

Experimen tal — Colorimetric determinations

XVIII, 10 Some general remarks upon colorimetric determinations
Cations

XVIII, 1 1 Aluminium

XVIII, 12 Determination of ammonia

XVIII, 13 Antimony

XVIII, 14 Arsenic

XVIII, 15 Beryllium

XVIII, 16 Bismuth

XVIIl, 17 Boron

XVIII, 18 Chromium

693

695

699

701

704

705
707

713

720

727

729

730

731
731
735

735

736
738

XXI

XVIII, 19
XVIII, 20
XVI1I,2I
XVIII, 22
xvin,23
XVIII, 24
XVIII, 25
XVIII, 26
XVIII, 27
XVIII, 28
XVIII, 29
XVIII, 30

Cobalt

Copper

Iron

Lead

Magnesium

Manganese

Molybdenum

Nickel

Tin

Titanium

Tungsten

Vanadium

Anions

XVIII, 31 Chloride
XVIII, 32 Fluoride
XVIII, 33 Nitrite
XVIII, 34 Phosphate
XVIII, 35 Silicate
XVIII, 36 Sulphate

Experimental — determinations with ultraviolet, -visihle
spectrophotometers

XVIII, 37 Dclcnnination of the absorption curve and concentration
of a substance (potassium nitrate)

XVIII, 38 Spcctrophotomctric determination of the pA' value of an
indicator (the acid dissociation constant of methyl red)
XVIII, 39 Simultaneous spcctrophotomctric determination
(chromium and manganese)

Experimental - determinations by spectrophotomelric titration
XVIII, 40 Spcctrophotomctric titrations
XVIII, 41 Apparatus for spcctrophotomctric titrations
XVni, 42 Simultaneous determination of arscnic{lU) and
antimony(Ill) in a mixture
XVIII, 43 Determination of coppcr(ll) with F.DTA
XVIII, 44 Determination of iron(lll) with FDTA
XVIII, 45 Determination of nickel ion with FDTA
XVIII, 46 References
XVIII, 47 Selected bibliography

739

740

741
744
744

746

747

748

750

751

752

753
753
755

755

756

757

758

759

761

763

767

768

769

770

771

772

Chapter XIX Fluorimetr)’ 773

XIX, I General discussion 773

XIX, 2 Instruments for nuorimctric analysis ^ ^ 774

XIX, 3 Some applications of Huorimciry 776

Experimental 777

XIX, 4 Quinine 777

xxii

XIX, 5

Aluminium

777

XIX, 6

Cadmium

778

XIX, 7

Calcium

779

XIX, 8

Zinc

779

XIX, 9

References

780

XIX, 10

Selected bibliography

780

Chapter XX Nephelometry and turbidimetry

781

XX, 1

General discussion

781

XX, 2

Instruments for nephelometry and turbidimetry

782

Some nephelometric determinations

784

XX, 3

XX, 4

Sulphate yy

Phosphate

784

785

XX, 5

Selected bibliography

786

Chapter XXI Emission spectrography

787

XXI, 1

General discussion

787

XXI, 2

Equipment for emission spectrographic analysis

789

XXI, 3

Qualitative spectrographic analysis

795

XXI, 4

Quantitative spectrographic analysis

798

XXI, 5

Direct reading instruments

800

Experimental

801

XXI, 6

QuaUtative spectrographic analysis of (a) a non-ferrous
alloy and (b) a complex inorganic mixture

801

XXI, 7

Determination of lead in brass ‘

803

XXI, 8

Determination of copper and lead in white metal

807

XXI, 9

Selected bibliography

808

Chapter XXQ Flame spectrometry

810

XXII, 1

General discussion

810

XXII, 2

Elementary theory

811

XXII, 3

Instrumentation

814

XXII, 4

Combustion flames

814

XXII, 5

The nebuliser-bumer system

Non-flame techniques

815

XXII, 6

817

XXII, 7

Resonance line sources

819

XXII, 8

Monochromator

820

XXII, 9

Detectors

,820

xxn, 10

Interferences

821

XXII, 11

Chemic'hi'interferences ,

822

Commercially available instruments

824

XXII, 12

Flame photometers

824

XXII, 13

Single beam atomic absorption spectrophotometers

' '826

xxiii

XXH. 14 Double beam atomic absorption spectrophotometers
XXII, 15 Atomic fluorescence spectroscopy

Experimental

XXll, 16 Evaluation methods
XXII, 17 Preparation of sample solutions
XXII. 18 Preparation of standard solutions
XXII. 19 Safety practices

Some selected (letcrmimitions
XXII. 20 Iniroduclion

XXII, 21 Experiments with a simple flame photometer
XXII, 22 Determination of magnesium and calcium in tap water
(AAS)

XXII, 23 Determination of vanadium in lubricating oil( AASI
XXII. 24 Detennination of trace lead in a ferrous alloy (.-NAS)
XXII, 25 Determination of chromium in a nickel alloy! A AS)
XXII, 26 Determination of sulphate ion by .Atomic Absorption
InhibitionTilrimetry
XXII, 27 References
XXII, 28 Selected bibliography

827

830

830
8.30

831

8.32

8.33

834

8.35

837

840

842

843

844

845

PARTG THERMAL METHODS

Oiapter XXIII Thermal analy.st.s 849

XXin, 1 General discussion 849

XX11I,2 Thermogravimetry (TG) 849

XXIII, 3 Instrumentation for thermogravimetry , 853

XX1II,4 Applications of thermogravimetry 855

XXIII, 5 Experimental 857

XXlll, 6 Differential Thermal Analysis and Differential Scanning

Calorimetry 859

XXIII, 7 Instrumentation for DTA and DSC 860

XXIII, 8 Experimental and instrumental factors 861

XXIII, 9 Application of Differential Thermal Analysis and

Differentia! Scanning Calorimetry 862

XXIII. 10 Experimental 862

Tltcrmometric titratioivs 864

XXIII, 1 1 Introduction 864

XXIII, 12 Theory gf ,5

XXIII, 13 Instrumentation 866

XXIII, 14 Applications 866

XXIII, 15 Experimental 867

XXIII, 16 References ggg

XXIII, 17 Selected bibliography 868

xxtv

Appendices 869

Appendix I International atomic weights 1973 870

Appendix II Index of organic chemical reagents 871

Appendix III Specific gravities of acids at 20 °C 880

Appendix IV Specific gravities of alkaline solutions at 20 °C 881

Appendix V Data on the strength of aqueous solutions of the

common acids and of aqueous ammonia 88 1

Appendix VI Saturated solutions of some reagents at 20 °C 882

Appendix VII Solubilities of some inorganic compounds in water

at various temperatures 883

Appendix VIII Sources of analysed samples 885

Appendix IX Buffer solutions and secondary pH standards 886

Appendix X Approximate pH values of some common reagent

solutions at about room temperature 887

Appendix XI Dissociation constants of some acids in water at

25 °C 888

Appendix XII Potentials of the common reference electrodes 890

Appendix XIII Polarographic half-wave potentials 891

Appendix XIV Tables of arc ‘raies ultimes’ and persistent lines

for spectrographic analysis 892

Appendix XV Percentage points of the /-distribution 895

Appendix XVI F-distribution 896

Appendix XVII Percentage points of the ^^-distribution 897

Appendix XVIII Four-figure logarithms 898

Index 901

XXV

XXII, 14 Double beam aiomic absorption speciropbotomcters
XXIl, 1 5 Atomic fluorescence spectroscopy

Experimenlal

XXII, 16 Evaluation metbods
XXn, 17 Preparation of sample solutions

XXll, 18 Preparation ofstandard solutions

XXII, 19 Safely practices

Some selected determinations
XXII, 20 Introduction

XXH, 2! Experiments with a simple (lantc photometer
XXll 22 Determination of magnesium and calcium in tap water

(AAS)

XXII, 23 Determination of vanadium in lubricating oiifAAS)

XXII. 24 Determination of trace lead in a ferrous alloy (AAS)
XXII, 25 Determination of chromium in a nickel alloy/ AAS)

XXII, 26 Determination of sulphate ion by Atomic Absorinion

InhibitionTitrimetry
XXII. 27 References
XXII, 28 .Selected bibliography

PARTG THERMAL METHODS

Chapter XXIII Xltcrmal analysis 849

XXIll, I General discussion 849

XXIII, 2 Thennogravimetry (TG) 849

XXllI, 3 Instrumentation for ihcnnogravimeir}' - 853

XXIII, 4 Applications of thermogravimetry 855

XXIII, 5 Experimental 857

XXIII, 6 Differential Thermal Analysis and Ditferenlial Scanning

Calorimetry' 859

XXIII, 7 Instrumentation for DTa\ and DSC 860

XXIII, 8 Experimental .and instrumental factors 861

XXIII, 9 Application of Diflcrenlial Thermal Analysis and

Dificrcntial Scanning Calorimetry 862

XXIII, 10 Experimental 862

Thcrmometric titrations 864

XXIII, 1! Introduction 864

XXIII, 12 Theory g^5

XXIII, 13 Instrumentation 866

XXIII, 14 Applications 866

XXIII, 15 Experimental 867

XXIII, 16 References 868

XXIII, 17 Selected bibliography 868

82?

830

831

832

833

834

835

837

840

842

843

844

845

XXIV

Appendices 869

Appendix I International atomic weights 1973 870

Appendix II Index of organic chemical reagents 87 1

Appendix III Specific gravities of acids at 20 °C 880

Appendix IV Specific gravities of alkaline solutions at 20 °C 881

Appendix V Data on the strength of aqueous solutions of the

common acids and of aqueous ammonia 88 1

Appendix VI Saturated solutions of some reagents at 20 °C 882

Appendix VII Solubilities of some inorganic compounds in water

at various temperatures 883

Appendix VIII Sources of analysed samples 885

Appendix IX Buffer solutions and secondary pH standards 886

Appendix X Approximate pH values of some common reagent

solutions at about room temperature 887

Appendix XI Dissociation constants of some acids in water at

25 °C 888

Appendix XII Potentials of the common reference electrodes 890

Appendix XIII Polarographic half-wave potentials 891

Appendix XIV Tables of arc ‘raies ultimes’ and persistent lines

for spectrographic analysis 892

Appendix XV Percentage points of the /-distribution 895

Appendix XVI F-distribution 896

Appendix XVII Percentage points of the ^^-distribution 897

Appendix XVIII Four-figure logarithms 898

Index 901

XXV

XXn, 14 Double beam aiomic absorption spcctropliolomelers
XXII. 1 5 Atomic fluorescence spectroscopy

Experimental

XXII. 16 Evaluation methods
XXII, 17 Preparation ol'samplc solutions

XXII, 18 Preparation of standard solutions

XXIl. 19 Safely practices

Some selected determimttions
XXII, 20 Introduction

XXll, 21 Experiments with a .simple flame photometer
XXII, 22 Determination of magnesium and calcium in tap water
(AAS)

XXII, 23 Determination of vanadium in lubricating oiI(y\AS)
XXII, 24 Determination of trace lead in a ferrous alloy (.AAS)

XXII, 25 Determination of chromium in a nickel alloy (AAS)

XXII. 26 Determination of sulphate ion by Atomic Ab.sorption
InhibilionTitrimetry
XXIl, 27 References
XXII, 28 Selected bibliography

827

830

831

832

833

834

835

837

840

842

843

844

845

PARTG

THERMAL METHODS

Chapter XXHI TIterma! analysis

849

XXIII, 1

General discussion

849

XX1II.2

Tiiennogravimctry (TG)

849

XX111.3

Instrumentation for themiogravimetry ^

853

XX1I1,4

Applications of Ihcnnogravimctry

855

XXIII. 5

Experimental

857

XXIII, 6

DilTercntiai Thermal Analysis and Diflercntial Scanning

Calorimetry

859

XXIII, 7

Instrumentation for DTA and DSC

860

XXIII, 8

Experimental and instrumental factors

861

XXIII, 9

Application of Differential Thermal Analysis and

Differential Scanning Calorimetry

862

XXIII, 10

Experimental

862

niermometric titrations

864

XXIII, 11

Introduction

864

XXIII. 12

Theory

865

XXIII, 13

Instrumentation

866

XXIII, 14

Applications

866

XXIII, 15

Experimental

867

XXIII, 16

References

868

XXIII, 17

Selected bibliography

868

XXIV

Appendices 869

Appendix I International atomic weights 1973 870

Appendix II Index of organic chemical reagents 87 1

Appendix III Specific gravities of acids at 20 °C 880

Appendix IV Specific gravities of alkaline solutions at 20 °C 881

Appendix V Data on the strength of aqueous solutions of the

common acids and of aqueous ammonia 88 1

Appendix VI Saturated solutions of some reagents at 20 °C 882

Appendix VII Solubilities of some inorganic compounds in water

at various temperatures 883

Appendix VIII Sources of analysed samples 885

Appendix IX Buffer solutions and secondary pH standards 886

Appendix X Approximate pH values of some common reagent .

solutions at about room temperature 887

Appendix XI Dissociation constants of some acids in water at

25 °C 888

Appendix XII Potentials of the common reference electrodes 890

Appendix XIII Polarographic half-wave potentials 891

Appendix XIV Tables of arc ‘raies ultimes’ and persistent lines

for spectrographic analysis 892

Appendix XV Percentage points of the /-distribution 895

Appendix XVI F-distribution 896

Appendix XVII Percentage points of the ^^-distribution 897

Appendix XVIII Four-figure logarithms 898

Index 901

XXV

FOREWORD

SI units have been used throughout this book, but with the acceptance of ‘litre’ as
a special name for the cubic decimetre we have the introduction of a non-SI term.

In this book the recommended convention has been adopted, namely that
concentration data of high precision are expressed in terms of the dm^, and only
data of moderate accuracy are expressed in terms of the litre.

Concentrations of solutions are usually expressed in terms of moles per cubic
decimetre : a molar solution (M) has one mole of solute per dm^.

For some purposes, however, it is more convenient to work in terms of
equivalents rather than moles (see Chapter X) ; a normal solution (N) has one
equivalent of solute per dm^.

xxvn

PREFACE TO THE FOURTH EDITION

Successive editions of Arthur I. Vogel’s books, and especially his Textbook of
Quantitative Inorganic Analysis, have become accepted internationally as
standard texts in colleges and laboratories. Only his untimely death in 1966
prevented him from extending and improving the various volumes which have
become so familiar to generations of undergraduates.

In carrying out the revision necessary for this Fourth Edition we have been
very conscious of the fact that we have been revising a book which possesses the
character of one particular person. Because of this we have sought to retain that
character throughout the reorganisation and introduction of new material. At
the same time we have made a number of changes which are intended to make the
individual sections and chapters as self-contained as possible. We have also
chosen to emphasise more strongly the importance of statistics and sampling to
the analytical chemist and have introduced a chapter on thermal analysis.

Some of the traditional methods of analysis which occupied substantial
sections of earlier editions have received fairly heavy pruning in order to create
sufficient space for the enlargement ,and introduction of sections dealing with
instrumental methods. It has been difficult to decide what to delete and what to
retain, but in those areas in which the third edition devoted space to several
titrimetric or gravimetric procedures for individual elements we have reduced the
number of entries to those of widest application.

The chapter on flame photometry has been rewritten to allow for the inclusion
of a more substantial section dealing with the development of atomic absorption
spectroscopy. At the same time the other chapters on spectroanalytical methods
have been substantially reorganised and extended. The chapter on infrared
spectrophotometry that was included in the third edition has been deleted in view
of ffie limited application of this technique for quantitative inorganic analysis.

The opportunity has been taken to rearrange the chapters dealing with
electroanalytical methods, and polarography is now included under the wider
name voltammetry’. We have also extended the section on separative techniques
0 include a chapter on gas chromatography because of its general application in
ana ytical chemistry, and in the context of this book for its use with certain
vomtile inorganic compounds.

d 1 important data in the appendices have been retained, but we have

th t rti chemical factors and the five figure logarithms as we have found

a ese are rarely used. A table giving a wide range of reagents suitable for the
e ermination of metals has been introduced (by permission of Hopkin and

XXIX

Williams Ltd.) listing a range of reagents suitable for determinations of metals.
Where necessary the other tables have also been rcvi.scd.

The work now occupies twenty-three chapters divided into seven parts. In this
edition we have included a number of references to the sources of the new
material that has been incorporated into the book. SI units have been employed
throughout and the chemical nomenclature up-dated.

We should at this stage like to express our sincere appreciation to the many
companies and publishers who have been so tolerant and understanding in
providing us with information, diagrams and photographs for the new edition.
That we had more thanVe could use is an indication of the great interest they
have taken in this production.

We are all indebted to our wives in many different ways for their
encouragement and assistance during the many months we have spent writing,
revising, checking and finally proof reading. That we have actually finished the
task is greatly due to their help throughout. Acknowledgement is also made of
the helpful discussions with colleagues and of the assistance given by members of
the laboratory staff of the School of Chemistry.

In conclusion we would like to say how pleased w-c are that we have been given
this opportunity to carry on the work of a man who did so much to promote high
standards in analytical chemistry. We hope that our efforts in producing the
Fourth Edition of this book will in themselves serve as part of the memorial to the
work of Arthur 1. Vogel.

J. Bassett R. C. Denney G. H. Jeffery J. Mendhara
Thames Polyiechnic. Woolwich, I^ndoti, S.E.IS.

PREFACE TO FIRST EDITIOW

In writing this book, the author had as his primary object the provision of a
complete up-to-date text-book of quantitative inorganic analysis,. both theory
and practice, at a moderate price to meet the requirements of University and
College students of all grades. It is believed that the material contained thereip is
sufficiently comprehensive to cover the syllabuses of all examinations in which
quantitative inorganic analysis plays a part. The elementary .student has been
provided for, and those sections devoted to his needs have, been treated in
considerable detail. The volume should therefore be of value to the student
throughout the whole of his career. The book will be suitable inter alia Jot
students preparing for, the various Intermediate B.Sc. and Higher School
Certificate Examinations, the Ordinary and Higher National Certificates in
Chemistry, the Honours and Special B.Sc. of the Universities, the Associateship
of the Institute of Chemistry, and other examinations of equivalent standard. It is
hoped, also, that the wide range of subjects discussed within its covers will result
in the volume having a special appeal to practising analytical chemists and to all
those workers in industry and research who have occasion to utilise methods of
inorganic quantitative analysis.

The kind reception accorded to the author’s Text Book of Qualitative Chemical
Analysis by teachers and reviewers seems to indicate that the general arrange-
ment of that book has met with approval. The companion volume on
Quantitative Inorganic Analysis follows essentially similar lines. Chapter I is
devoted to the theoretical basis of quantitative inorganic analysis. Chapter II to
the experimental technique of quantitative analysis. Chapter III to volumetric
analysis, Chapter IV to gravimetric analysis (including electro-analysis). Chapter
V to colorimetric analysis, and Chapter VI to gas analysis; a comprehensive
Appendix has been added, which contains much useful matter for the practising
analytical chemist. The experimental side is based essentially upon the writer’s
experience with large classes of students of various grades. Most of the
determinations have been tested out in the laboratory in collaboration with the
author’s colleagues and senior students, and in some cases this has resulted in
slight rnodifications of the details given by the original authors. Particular
emphasis has been laid upon recent developments in experimental technique.
Frequently the source of certain apparatus or chemicals has been given in the
text ; this is not intended to convey the impression that these materials cannot be
obtained from other sources, but merely to indicate that the author’s own

experience is confined to the particular products. mentioned.

xxxi

FUNDAMENTALS
OF QUANTITATIVE
PART A INORGANIC ANALYSIS

■ The ground covered by the hook can best be judged by perusal ol (lie Rible of
Contents. An attempt has been made to strike a balance between thcclassiral and
modern proccdurc.s'. and to present the subject of analytical chemistry as it is to-
day. The theoretical aspect htis been stressed throughout, and numcrmis cross-
references arc given to Chapter 1 (the theoretical basis of quantittitivc inorganic
analysis).

No references to itic origimd literature arc given in the te.Nt. This is because the
introduction of sucli references would have considerably increased the size and
therefore the price of the book. However, a discussion on the literature of
analytical chemistry is given in the Appendix f Section A, 3). With the aid of the
various volumes mentioned therein — which should be available in all libraries of
analytical chemistry— and the Collective Indexes of Chemical Ahitrans or of
British Chemical Abstracts, lilllc difTiculty will, in general, be experienced in
finding the original sources of most of the detcmiinations described in the book.

In the preparation of this volume, the author has utilised pertinent material
wherever it was to be found. While it is impossible to acknowledge every' stiurcc
individually (sec, forcxnmpic. Section A. 3), mention must, however, be made of
Hillcbrand and Lundcli'.s Applied htorpanic Arta/vsis (\‘)29} and of Mitchell and
Ward’s Modern Methods in Quantitative Chrmieal Analysis (1932). In con-
clu-sion. the writer wishes to express his thanks : to Dr. G. H. Jefiery, A.I.C., for
reading the galley proofs and making numerous helpful stiggesiions; to Mr. A, S,
Nickclson, B.Sc,. for reading some of the galley proofs; to his laboratory
steward. Mr. F. Mathie, forpreparinga numberof thetliagrams. including most
of those in Chapter VI, and for his assistance in other ways; to Messrs. A.
Gtillcnkantp and Co,. Ltd,, of London, E.C.2. and to Messrs. Fisher Scientific
Co, of Pittsburgh, Pa., for providing a number of diagrams and blocks;* and lo
Mr. F. W. Clifford, F.L.A., Librarian to the Chemical Society, and his able
assistants for their help in the task of searching the extensive literature.

Any suggestions for improving the book will be gratefully received bv the
author.

Woolwich Polytechnie. Lamkm. S.C.IS. June. 1939.

tcnowlcdgmcnl to olhtr firm'i anti intliv'nJiuils, made in ilic body of the text.

xxxii

FUNDAMENTALS
OF QUANTITATIVE
INORGANIC ANALYSIS

■ The ground covered by the book can best be judged by perusal olThc Table of
Contents. An attempt has been made to strike a balance between the classical and
modern procedures, and to prc.scnt the subject of analytical chemistry as it is to-
day. The theoretical aspect has been stressed throughout, and numerous cross-
references arc given to Chapter I (the theoretical basis of quantitistivc inorganic-
analysis).

No references to the original literature are given in the text. This is because the
introduction of such references would have considerably increased the si?c and
therefore the price of the book, However, a discussion on the literature of
analytical chemistry is given in the Appendix (Section A, 3). With the aid of the
various volumes mentioned therein - which should be available in all libraries of
analytical chemistry -and the Collective indc.xes of Clivniiral Ah\iriicis or of
Briiish Chemical Ahstracis, little dinkulty will, in general, be experienced in
finding the original sources of most of the determinations described in the book.

In the preparation of this volume, the author has utilised pertinent material
wherever it was to be found. While it is impossible to acknowledge every source
individually (see. for example. Section A, 3), mention must, however, be made of
Hillcbrand and Lundell's Applied Iimrpiinir Amdysi.% ( 1 929) and of Mitchell and
Ward’s Modern Method'^ in Qi/antiiatire Chemical Analysis (1932). In con-
clusion, the writer wishes to express his thanks; to Dr. G. H, Jeffery, A.I.C., for
reading the galley proofs and making numerous helpful suggestions; to Mr. A. S,
Nickclson, B.Sc.. for reading some of the galley proofs; to his laboratory
steward, Mr. F. Mathie, for preparing a number of the diagrams, including most
of those in Chapter VI. and for his a.ssistance in other way.s; to Messrs. A.
Gallenkamp and Co.. Ltd,, of London, E.C.2, and to .Messrs. Fisher SciciUtfic
Co, of Pittsburgh. Pa., for providing a number of diagrams and blocks;* and to
Mr. F. W. Clifford, F.L.A.. Librarian to the Chemic.'il Society, and his able
assistants for their help in the task of searching the extensive literature.

Any suggestions for improving the book will be graiefullv received by the
author.

Waolw ich Polytechnic, Limdun. S./C.J8. Jane. 19S9.

ledgmcniio other firms and individuals is made in Ihe body of the text.

xxxii

FUI\IDAiVIEi\ITALS
OF QUANTITATIVE
PART A INORGANICANALYSIS

CHAPTER I IWTRODUCTIOW

I, 1. ■ CHEMICAL ANALYSIS. ‘The resolution of a chemical compound
into its proximate or ultimate parts; the determination of its elements or of the
foreign substances it may contain’: thus reads a dictionary definition.

This definition outlines in very broad terms the scope of analytical chemistry.
When a completely unknown sample is presented to an analyst, the first
requirement is usually to ascertain what substances are present in it. This
fundamental problem may sometimes be encountered in the modified form of
deciding what impurities are present in a given sample, or perhaps of confirming
that certain specified impurities are absent. The solution of such problems lies
within the province of qualitative analysis and is outside the scope of the present
volume.

Having ascertained the nature of the constituents of a given sample, the analyst
is then frequently called upon to detennine how much of each component, or
of specified components, is present. Such determinations lie within the realm
of quantitative analysis, and to supply the required information a variety of
techniques is available.

I, 2. SAMPLING. The results obtained for the proportion of a certain
constituent in a given sample may form the basis of assessing the value of a large
consignment of the commodity from which the sample was drawn. In such cases it
is absolutely essential to be certain that the sample used for analysis is truly
representative of the whole. When dealing with a homogeneous liquid, sampling
presents few problems, but if the material under consideration is a solid mixture,
then it is necessary to combine a number of portions to ensure that a
representative sample is finally selected for analysis. The analyst must therefore
be acquainted with the normal standard sampling procedures employed for
different types of materials.'

I, 3. TYPES OF ANALYSIS. With an appropriate sample available,
attention must now be given to the question of the most suitable technique or
techniques to be employed for the required determinations. One of the major
decisions to be made by an analyst is the choice of the most effective procedure for
a given analysis, and in order to arrive at the correct decision, not only must he be
familiar with the practical details of the various techniques and of the theoretical
principles upon which they are based, he must also be conversant with' the
conditions under which each method is reliable, must be aware of possible

■ INTRODUCTION 1,5

Analytical Chemistry : Fresenius and }a.nder, Handhuch der analytischen Chemie ;
of a compendium of methods such as Meites,' Handbook of Analytical Chemistry;
or of specialised monographs dealing with particular techniques or types of
material. It may be necessary to seek for more up-to-date information than that
available in the books which have been consulted and this will necessitate making
net' (e.g. Annual Reports of the Chemical Society , Selected

Society for Analytical Chemistry), and of abstracts (e.g,
hemical Abstracts), and referring to journals devoted to
1 to specific techniques.*

vey may lead to the compilation of a list of possible
imate selection must then be made in the light of the
iciated, and with special consideration being given to
erferences and to the equipment available. '

HNIQUES. The main techniques employed in quanti-
s are based upon (a) the quantitative performance of
>ns and either measuring the amount of reagent needed
on, or ascertaining the amount of reaction product
te electrical measurements (e.g. potentiometry); (c) the
optical properties (e.g. absorption spectra); or (d), in some
f optical or electrical measurements and quantitative
...nperometric titration).

The quantitative execution of chemical reactions is the basis of the traditional
or ‘classical’ methods of chemical analysis: gravimetry, titrimetry and volumetry.
In gravimetric analysis the substance being determined is converted into an
insoluble precipitate which is collected and weighed, or in the special case of
electrogravimetry, electrolysis is carried out and the material deposited on one of
the electrodes is weighed.

In titrimetric analysis (hitherto often termed volumetric analysis), the
substance to be determined is allowed to react with an appropriate reagent added
as a standard solution, and the volume of solution needed for complete reaction is
determined. The common types of reaction which find use in titrimetry are (a)
neutralisation (acid-base) reactions; (b) complex-forming reactions; (c) pre-
cipitation reactions; (d) oxidation-reduction reactions.

Volumetry is concerned with measuring the volume of gas evolved or absorbed
in a chemical reaction.

Electrical methods of analysis (apart from electrogravimetry referred to above)
involve the measurement of current, voltage or resistance in relation to the
concentration of a certain species in solution. Techniques which can be included
under this general heading are (i) voltammetry (measurement of current at a
micro-electrode at a specified voltage); (ii) coulometry (measurerhent of current
and time needed to complete an electrochemical reaction or to generate sufficient
material to react completely with a specified reagent); (iii) potentiometry
(measurement of the potential of an electrode in equilibrium with an ion to be
determined); (iv) conductimetry (measurement of the electrical conductivity of a
solution).

Optical methods of analysis are dependent either upon (i) the absorption of

* Selected bibliographies are given in Section 1, 12 and at the conclusion of each chapter.

(jj

— ^ -“N IIBD

Iqpof ‘XjisMAinn •A'NT
rjqn jCpnaBa SauaanjSaa

1, 4 QUANTITATIVE INOFIGANIC ANALYSIS

interferences wliich tnay arise, and must be capable of devising means of
circumventing such problems. lie will also be concerned with quc.stions
rccarding the accuracy and the precision to l>e cxjKCtcd from giren mctliod.s and,
in addition, must not overlook such factor.s as tinte ,tnd costing. *Thc most
accurate method for a certain determination may prove to be lengthy or to
involve the use of expensive reagents, and in the interests of economy it may be
necessary to choose a method which, ailltough somenbat less exact, yields results
of sulUcient accuracy in a reasonable time.

Important factors which must be taken into account when .selecting an
appropriate method of analysis include (<j) the nature of the information which is
sought, (/)) the size of sample available and the proportion of the constituent to be
determined, and (c) the purpose for which the analytic.al data are required.

The nature of the information sought; this may consist of a requirement for
very detailed data, or alternatively, results of a general character may suffice.
With respect to the information which is furnisiied, different types of chemical
analysis may be classified as foliow.s:

fi) praximtirc aiuilyxis. in which the amount of each element in a sample is
determined with no concern as to the actual compounds present;

(ii) partial mutlysis, which deals with the determination of selected constituent.s
in the sample;

(iii) trace constituent analysis is a .specialised instance of partial analy.si.s in
which we arc concerned with the determination of specified components
prc-sent in very minute quantity;

(iv) complete analysis, when the proportion of each component of the sample is
determined.

On the basis of sample size, analytical methods arc often classified a.s:

Macro, the determination of quantities ofO.l g or more;

Semi-micro, dealing with quantities ranging from 0.01 g to 0.1 g;

Micro, for quantities not exceeding 0.001 g. The term semi-micro is not very
apt, referring as it docs to quantities larger than micro and it has been proposed
that it should be replaced by the term mesa.

A major constituent is one present in excess of 1 per cent, .a minor constituent is
one constituting from 0.0! to I per cent of the sample, and a trace const iliicnl is
one present to an extent of less than 0.01 per cent of the sample.

The purpose for which the analytical data arc required tnay perhaps be related
to process control and quality control. In such circumstances the objective is
checking that raw materials and finished products conform to spcdfication, and
It may also be concerned with monitoring various stages in a manufacturing
process. For this kind of determination, methods must be employed which are
quic and which can be readily adapted for routine work; in this area

important role to play, and in certain cases may
automation. On the other hand, the problem may be one
Which requires detailed consideration and which may be regarded as bcinc more
m the nature of a research topic.

anniuc^lwu LITERATURE. Faced with u research-type problem the
MDprifnw- “ situation which is outside his normal

inS 1 seek guidance from published data. This will

Flvinl of multi-volume reference works such as Kollhoff and

g, neatise on Analytical Chemistry, Wilson and Wilson, Comprehensive

INTRODUCTION 1,5

Analytical Chemisay; Fresenius and Jander, Handbuch der analytischen Chemie\
of a compendium of methods such as Meites, Handbook of Analytical Chemistry;
or of specialised monographs dealing with particular techniques or types of
material. It may be necessary to seek for more up-to-date information than that
available in the books which have been consulted and this will necessitate making

(e.g. Annual Reports of the Chemical Society; Selected

r =yvrE?foartcr Society for Analytical Chemistry), and of abstracts (e.g.

:hemical Abstracts), and referring to journals devoted to
d to specific techniques.* ■ ' ’

rvey may lead to the compilation of a list of possible
•imate selection must then be made in the light of the
nciated, and with special consideration being given to
ioqmv :erferences and to the equipment available.

Jqpor ‘XjisJaAinfi •A'NT

uq]'I SaiiaaniSns

IHNIQUES. The main techniques employed in quanti-
is are based upon (a) the quantitative performance of
ons and either measuring the amount of reagent needed
ion, or ascertaining the amount of reaction product
ite electrical measurements (e.g. potentiometry); (c) the
optical properties (e.g. absorption spectra); or (d), in some
if optical or electrical measurements and quantitative
mperometric titration).

The quantitative execution of chemical reactions is the basis of the traditional

or ‘classical’ methods of chemical analysis: gravimetry, titrimetry and volumetry.
In gravimetric analysis the substance being determined is converted into an

insoluble precipitate which is collected and weighed, or in the special case of
electrogravimetry, electrolysis is carried out and the material deposited on one of
the electrodes is weighed. !

Volumetry is concerned with measuring the volume of gas evolved or absorbed
in a chemical reaction.

r*itliiint energy fiiul fi

^"<1 (V) />j/,-^L.,/ / (colorimetry) il,\’,ii, 'nvoivcdas

sSf "’^ -v

1“'.Tmi„„,. •'" "K'ric lamp proJ,,d„E V “ ^ 'I'n’',' '/?

=tn“tr; "— • '■" »«d. ,,„ ,

, may uUo be

..r dcdH.,

'‘I ^'■"■/'““'"•'n'.mwhicl,. , "’"’''"‘''''l■'■■»''••' lesion) U

Si?""

wamined. "' charnctcriS. 'WtJi visible c

emitted radiation is the

n'TUr^r^

I. 6 . OTHER TECHN ''

S-^ral methods o

‘o replace the In« i “'‘^‘^'’’on from th,> ^-’’ays ofshort u--., . ^ 's
shells and in '''miher dp "*^'‘'"«st electron shell of '

‘secondary or may jump ?rom o°l‘'’"‘'''°"’- -'nd

characteristic of the ^’'‘‘''y radhr o'" X-rayr o«ler
'o assess theamoimro"nh"'T'"‘'‘‘^’‘'‘"d (hS ‘^"’'''ted at^

? :m example of a numi "'^^'‘'ment giving "‘'^"sity ofthc radiation!^ .‘■'"S''is
he outside the ce,d ™^‘^''®^so-ca||ed nr, ® '* Present in ^c used

"oniins.nc ?vV?S' °'' too" A m,'"?i"''i™So* ‘'’‘■"'"Pl^ 71, «

*'^“"'■'2 “““ ""'■' toe„?„">'or,ox,sSw S^^

'srp's S“ r of?s^s

6 "'""hgation ,0 a

INTRODUCTION 1,7/8

Kinetic methods of quantitative analysis are based upon the fact that the speed
of a given chemical reaction may frequently be increased by the addition of a
small amount of a catalyst, and within limits, the rate of the catalysed reaction
will be governed by the amount of the catalyst present. If a calibration curve is
prepared showing variation of reaction rate with amount of catalyst used, then
measurement of .reactioii rate will make it possible to determine how much
catalyst has been added in a certain instance, and this provides a sensitive method

for determining sub-microgram amounts of appropriate substances. ■ , :

I, 7. INSTRUMENTAL METHODS. The methods dependent upon meas-
urement of an electrical property, and those based upon -determination of
the extent to which radiation is absorbed or upon assessment of the intensity of
emitted radiation, all require the use of a suitable instrument, e.g. polarograph,
spectrophotometer, etc., and in consequence such methods. are referred to as
instrumental methods. Instrumental methods are usually much faster than purely
chemical procedures, are normally applicable at concentrations far too small to
be amenable to determination by classical methods, and find wide application in
industry. In many cases a recorder can be attached to the instrument so that
absorption curves, polarograms, titration curves, etc., can be plotted automati-
cally, and in fact, by the incorporation of appropriate servo-mechanisms, the
whole analytical process may, in suitable cases, be completely automated.

Despite the advantages possessed by instrumental methods in many directions,
their widespread adoption has not rendered the purely chemical or .‘classical’
methods obsolete; the situation is influenced by three main factors.

1. The apparatus required for classical procedures is cheap and readily available
in all laboratories, but many instruments are expensive and their use will only
be justified if numerous samples have to be analysed.

2. With most instrumental methods it is necessary to carry out a calibration

operation using a sample of material of known composition as reference
substance; the exact analytical data for this standard must be established by
alternative procedures which will normally mean by the use of classical
chemical methods. ,

3. Whilst an instrumental method is ideally suited to the perforihance of a large
number of routine determinations, for an occasional, non-routine analysis, it is
often simpler to use a classical method than to go to the trouble of preparing
requisite standards and carrying out the calibration of an instrument.

Clearly, instrumental and classical methods must be regarded as supplement-
ing each other. ,

1,8. TIME AND MONEY, ACCURACY AND RANGE. Salient infor-
mation relating to the more common quantitative.techniques is presented below;
The methods are arranged in columns according to the cost of the main
equipment involved:

Cheap, lessthanflOO; '

Moderate, in therange£200-£1000; ' '

High, in the range £1000-£4000;

Expensive, in excess of £4000.

This division is obviously somewhat arbitrary, and the cost of an instrument for a
given techmque vanes widely according to the degree of sophistication offered;.

1, 9 QUANTITATIVE INORQANIC ANALYSIS

The columns have been diviiicd to give an indication of the lime required for
each technique. At lire head ofcacli column arc the ‘fa.sf procedures, that is. those
wliich can normally be completed in ten to fifteen minutes, whilst the methods
grouped at liie foot of each column arc the ‘slow' procedures, those which
normally require more than one hour for completion. Speed of operation is \cr\
much a personal factor, and it must be ctnpbasised that the time required for
preparing the final solution for analysis (including the removal of interferences)
has not been taken into account, nor has the time needed for instrument
calibration or for standardisation procedures.

Other information included is an indication by means of the lettcr.s H, M, L of
the accuracy normally to be expected from a given method in the hand.s of a
competent analyst:

1! high accuracy, results Iselter than 1 per cent;

M moderate accuracy, results in the range 1 per cent- S per cent :

L low accuracy, results not better than 5 per cent.

The numeral following the letter showing liie accuracy provides an indication
of the concentration range in which the method can be satisfactorily employed;

1 g—cg per cubic decimetre;

2 g — mg per cubic decimetre;

3 g~-pg per cubic decimetre;

4 dg—mg per cubic decimetre;

5 mg— /(g per cubic decimetre;

6 g— mg actual sveight;

7 dg—mg actual weight (macro analysis);

8 p.p.m.— p.p.b.;

9 1 percent — 0.(X31 percent;

10 1 per cent — p.p.b.

I, 9. INTERFERENCES, Whatever the method finally chosen for the
required determination, it should, ideally, be a .spedfir mciltoil; that is to say, it
should be ctipable of measuring the amount of desired substance accurately, no
matter what other substances may be pre.scnt. In practice few analytical
procedures attain this ideal, but many nictliods are selective; in other words, they
can be used to determine any of a small group of ions in the pre.scncc of certain
specified ions. In many instances the desired selectivity is achieved bv carrx’ing
out the procedure under carefully controlled conditions, particularly with
reference to the pH of the solution.

Frequently, however, there arc substances present that prevent direct
measurement of the amount of a given ion ; these arc referred to as interferences,
and the selccUon of methods for separating the interferences from the substance
to be determined are as important as the choice of the method ordclcrminntion.
Typical separation procedures include the follow-ing:

(a) Selective precipitation. The addition of appropri.atc reagents may convert
interfering ions into precipitates which can be filtered off, careful pH control is
often necessary in order to achieve a clean separation, and it must be borne in
mind that precipitates tend to adsorb substances from solution and care must be
taken to ensure that as little as possible of the substance to be determined is lost in
this way.

(b) Masking. A complexing agent is added, and if the resultant complexes are
sufficiently stable they will fail to react with reagents added in a subsequent

'* ^^UANTITATIVK INORgaKJC AKAI.VSI.S

The columns have been . i .

SiiiilP

iliSl5?====---

5 me—/ n decirneirc-

7 ''"■''SfX;

* p.p.m.-fp’p'b'!' “'*''' •"’•■"'o o'wlysid:

« “■= °" c V,;; ;;;« 'j “o™„

'"<':rtcri„’cicl:J''"'<’''‘''‘>’’i- TlS'S '“"““mg: ""' "^''"'■™'iiialion

S S;f ? ''"'pXtS” “■'’-h-"!- ™»y convc

mind that nr • '" lo achL ’ filtered off "r"'*'' ‘-’O"'''-''

fi'>'^en o^P'‘‘'>^ tend totr*^ " separa, bn n ’ P'^ i

this wir^“^'^''>a,as,itdeS;\t4 ■'

Maskh, A ° substance tor,.. fi'

3 ^“"’Plc.xing anen, ..... ‘^<='crmined is Jost ir

Baskin A ' •'^'‘’'‘^“^‘'’‘-'substanco.ob t

^ffficientlyV fbf ''?"’P'e.xing agent f . ''s Jost i

I- >0. J1

pliOspiiaics is <|((]' I

‘^‘"'unin.-ildiiTcrcntrih- or.-, voL,'!,!' -.r V » scpnrajion

termed (he .-•,/• beini' n n'i .,1 ’ -^s ^‘■“■’ P'-i'cl do«n -,

'O flic top of the L?, >'‘rccmp|o\c«! H-.. m.iicrials as adlulov

"»«■ ”s- "f '« 'It:

'^oniponcnis arc ehitJ ’ ‘.'hitcd hrvt •,„.( f, '•^' '"‘^*’'1^' pfiasc; ih;

slouiy. ,f,„, "'f'fc-rcadiiy adsorbed

liquid (coinnionlv ii iv,.|i. I In piinHwn

o:i^r"’'^nhaseaml:;S

^oncentrated hydrochlorV V ’V''-'''''’'’'''"’'' ‘'fcof, ^tationarv

" f'fe Phase, acetone CO , '''■‘‘'-nart th ' in

‘‘ shceVof er '>'froj.'cn. It fs

;iasE !s s;£?£==#H5H

in tn^ T” nios.

‘^^or of^hSir'"'^ "'-'P^rf 'ami'’, f^ii/i'''then

tj . Thcanalvt •ii^“"''''”Pfit>csn(,i;. ? ^''■’ ‘•‘''‘■'lulled to

value for iiie''' *'*“ "itli ii,e ,.1^" ‘‘ ''‘‘•'f of results for tb,^ question of

S ImpSt™ '!» follo«i„5 Slop, ,r

5 i‘aiSiSa~

INTRODUCTION I, 12

I, 12. Selected bibliography ...

The following selection of reference books, journals, and review articles is not
intended to be exhaustive. The books listed include the widely known general
textbooks and reference works devoted to quantitative chemical analysis, and to
books giving general accounts of various topics referred to in this chapter.
Succeeding chapters carry individual bibliographies which include the more
important books dealing with specialised techniques.

A. General reference books

1 . ASTM Standards. American Society for Testing Materials, Philadelphia, 1 964.

2. D. Abbott and R. S. Andrews (1970). An Introduction to Chromatography, 2nd edn.
London ; Longman.

3. J. A. Barnard and A. Chayen (1965). Modern Methods of Chemical Analysis. London;
McGraw-Hill.

4. R. Belcher and A. J. Nutten (1970). Quantitative Inorganic Analysis. 3rd , edn.
London; Butterworth.

5. R. Belcher and C. L. Wilson (1964). New Methods of Analytical Chemistry. 2nd edn.
London; Chapman and Hall.

6 E. W. Berg (1963). Physical and Chemical Methods of Separation. New York;
McGraw-Hill. . ' , •

7. W. G. Berl (1960). Physical Methods in Chemical Analysis. 2nd edn. New York;
Academic Press.

8. D. Betteridge and H. E. Hallam (1972). Modern Analytical Methods. London; The
Chemical Society.

9. D. R. Browning (1969). Electrometric Methods. London ; McGraw-Hill; ■

10. G. Chariot (1960). Les methodes de la Chimie Analytique: analyse quantitative
minerale. 4th edn. Paris ; Masson et cie.

II. J. A. Dean (1969). Chemical Separation Methods. 'lAtvi York; Van Nostrand.

12. G. W. Ewing (1968). Instrumental Methods of Chemical Analysis. 3rd edn. New York;
McGraw-Hill.

13. G. W. Ewing (1971). Topics in Chemical Instrumentation. Easton ; Chemical Education
Pub. Co.

14. W. Fresenius and G. Jander (from 1944). Handbuch der Analytischen Chemie, Dritter

TeiY. Berlin; Springer-Verlag. . . '

15. N. H. Furman and F. G. Welcher(1962). Standard Methods of Chemical Analysis. 6th
edn. Princeton; Van Nostrand.

16. W. F. Hildebrand, G. E. J. Lundell, H. A. Bright and J.'A. Hoffman {X^ST). Applied

/norgan/c 2nd edn. New York; Wiley.

17. K. Kodoma (1963). Methods of Quantitative Inorganic Analysis: an encyclopaedia of
gravimetric, titrimetric and colorimetric methods. XAev/YoTkiWiley. .

18. I. M. Kolthoff and P. J. Elving (from 1959). Treatise on Analytical Chemistry. New

York; Wiley. ■ . ,

19. P. Kruger (1971). Principles of Activation Analysis. New York; Wiley-Interscience. '

20. N. A. Lange (1966). Handbook of Chemistry. 10th edn. New York; McGraw-Hill.

21. H. A. Liebhafsky, H. G. Pfeiffer, E. H. Winslow, P. D. Zemany, and S. S. Liebhafsky

(1972). X-ray Absorption and Emission in Analytical Chemistry. New.York; Wiley-
Interscience, • ' ■ . . .

22. L. Meites (1963). Handbook of Analytical Chemistry. New York; McGraw-Hill.

23. W. F. Pickering (1971). Modern Analytical Chemistry. fXew York; Mar^djkkker.

W'f (1963). Ion-exchange Separations in Analytical Clientisl^^^&Y ovk\

25. A. Seidell and W. F. Linke (1958). Solubilities of .'Inorganic

4th edn. Princeton, Van Nostrand. ' " : ; .

I, 12 QUANTITATIVE INORGANIC ANALYSIS

">6 S Sigcia(!968). Survey oj AnalytimI Chemistry. New York; McGraw-Hill.

2?! L. G.'Sillcn and A. e! Maricll (1964). StahiUiy Consimus of Maul Ion Complexes.
London: The Chemical Sodciy.

28. D. A. Skoogand D. M. Wcsi(1970). Fundumetuals of .■inahiical Chanhtry. 2ndcdn,
London: Holt. Rinehart and Winston.

29. D. A.Skoogand D, M. West (1971). Principles of Jnsiiiimeiitcil Aniily.m. New York:
Holt, Rinehart and Winston.

30. F. D. Snell and C. L. Hilton (from 1969). I'.ncyclopaetlin <f Industrial Chemical
Analv.vis. New York: W'ilcy.

31. C. R. N. Strouts. H. N. Wilson and R.T. R.irry-Jo!ies(I962). Chemical Analysis: the
work-ins; look. 2nd edn. O.vford ; O.xford Universits Press,

32. R. C. Weast (1972). Handbook of Chemistry and Physic.y 53rd cdn. Cleveland;
Chemical Rubber Publishing Co.

33. R. K. Webster (I960). Methods in Geochemistry. London; Inlerscicnce.

34. F. .1. Wclchcr (1947). Organic Analytical Reagents. Princeton : Van Noslratid.

35. T. S. West (1972). Analytical Chemistry t Parts I. 2). MTP Series 1, Vols. 12. 13.
London; Ruitersvorths.

36. H. H. Willard. L. L, Merritt and J. A. Dean (197-1). Instrumental Methods of Analysis.
5th cdn. New York; Van Nosirand.

37. C. L. Wilson and D. W, Wilson (from 1959). Comprehensive Analytical Chemistry.
Amsterdam; Elsevier.

38. J. G. Dick(1973). .-Inulytical Chemistry. New York; McGraw-Hill Book Co.

39. J. S. Fritz and G. H. .Schtnek (1974). Quantitative Analytical Chemistry. 3rd cdn.
Boston; Allyn and Bacon.

40. D. G. Peters. J. M. Hayes and G. M. HieRje (1974). Chemical Separations and
Measurements. Philadelphiti; W. B. Saunders Co.

41. W. E. Harris and B. Kralochvil (1974). Chemical Separations and Mea.suremcnts.
Philadelphia; W, B. Saunders Co.

(This is complementary to No. 40.)

42. H, A. Strobel (1973). Chemical Instrumentation — A Systematic Approach to
Instrumental Analysis. 2nd cdn. Reading, Mass. ; Addison-Wesley Publishing Co.

43. D. J. Pietrzyk and C. W. Frank (1974). Analytical Chemistry: An Introduction. New
York; Academic Press Inc.

44. H. F. Walton and J. Reyes (1973). Modern Chemical .-inaly.sis and Instrumentation.
New York; Marcel Dckkcr Inc.

45. T. H. Gou\v(I972). Guide to Modern .Methods of Insinimcntal .Analysis. New ^'ork; J.
Wiley and Sons Inc.

46. J. W. Robinson (1973). Undergraduate Instrumental Analysis. 2nd cdn. New York;
Marcel Dekkcr Inc.

47. H. A. Laitincn and W. E. Harris (1975). Chemical Analysis. 2nd cdn. New York;
McGraw-Hill.

48. p. de Soctc, R. Gijbcis and J. Hoslc (1972). Xeutron .Activation Analysis. Chichester,
Wiley.

so ^ (1971). Principles of Activation Analysis. New York; Wilcy-lnterscienec.

■ G. p. Chase and J. L. Rabinowii?. (1962). Principles of Radioisotope Methodology.
2nd edn, Minneapolis; Burgess.

S 7 r' ^ ? McKay (1971). Principles of Radiochemistry. London: Butterworths.

• • Anderson (cd.) (1973). Microprohe .4naly.sis. New York: Wiicv-lntcrsciencc.

^3. H A, Lrchhafsky, H. G. PfcilTcr, E. H. Winsiow and P. D. Zemany (1972). X-ravs,
t-Icctrons and Analytical Chemistry. Chichester; Wiley.

M Vries (1970). Practical A'-rar Spectrometry. London;

INTRODUCTION I, 12

B. Journals, abstracts and reviews

1 . Advances in Analytical Chemistry and Instrumentation.

2. Analytical Abstracts.

3. Analytica Chimica Acta.

4. Analytical Chemistry (includes Annual Review in April issue).

5. Annual Reports of the Chemical Society, London.

6. Chemical Abstracts.

7. Chemia Analityczna.

8. Chemical Titles.

9. Chimie Analytique.

10. Current Chemical Papers. ■

1 1 . Journal of Analytical Chemistry of the USSR ( Zhurnal analitischeskoi Khimii ) .

12. Selected Annual Reviews of the Analytical Sciences.

13. Spectrochimica Acta.

14. Talanta.

15. The Analyst.

16. Zeitschriftfur analytische chemie. ,

17. Journal of Electroanalytical Chemistry.

18. Journal of the Polargraphic Society.

19. Journal of Scientific Instruments.

20. Mikrochimica Acta.

FUWDAMEWTAL

CHAPTER II THEORETICAL PRiNCIPLES

II, 1 . ELECTROmiC DISSOCIATION. Many of the reactions of
qualitative and quantitative analysis take place in aqueous solution. Ii is
therefore necessary to have a genera! knowledge of the conditions which exist in
such solutions. It is assumed that the reader is familiar w'ith the broad concepts of
the simple theory of electrolytic dissociation.

Ionisation of acids and 1):lscs in solution. An acid may be defined as a
substance which, when dissolved in water, undergoes dissociation with the
formation of hydrogen ions as the only positive ions:

HCl ^erHM-Cl-
HNOj :?iir +NO,'

Actually the hydrogen ion H * (or proton) does not c.xist m tlie free state in
aqueous solution; each hydrogen ion combines with one molecule of water to
form the hydroxoninm ion HjO‘. The hydroxonium ion is a hydrated proton.
The above equations arc therefore more accurately written :

HCl + IliO;;e:HjO* +Cr
HN03 + Hj0=?iHj0"

Tlie ionisation may be attributed to tlic great tendency of the free hydrogen ions
to combine with water molecules to form hydroxonium ions. Hydrochloric
and nitric acids arc almost completely dissociated in aqueous solution in
accordance with the above equations; this is readily demonstrated by freexing-
point measurements and by other methods.

Polyprotic (polybasic) acids ioni.se in stapes. In sulphuric acid, one hydrogen
atom is almost completely ionised ;

+HSO4-

or + +HS04-

The second hydrogen atom is only partially ionisctl, except in very dilute

HS04-^iH"+S0^2-
or HSOr +H ,0 +S 0 /-

Phosphoric acid also ionises in stages;

H3PO4 +H2PO4' ^ 2 H'* J-HPO

^ 3 l-r +PO4I'

FUNDAMENTAL THEORETICAL PRINCIPLES H, 1

or H3P04+H20^H30++H2P04" , ,

H2PO4- +H2O-H3O+ +HPO42-

HP04"-+H20^H30++P04"-

The successive stages of ionisation are known as the primary, secondary, and
tertiary ionisations respectively. As already mentioned, these do not take place to
the same degree. The primary ionisation is always greater than the secondary,
and the secondary very much greater than the tertiary.

Acids of the type of acetic acid CH3COOH give an almost normal freezing-
point depression in aqueous solution; the extent of dissociation is accordingly
small. It is usual, therefore, to distinguish between acids which are completely or
almost completely ionised in solution and those which are only slightly ionised.
The former are termed strong acids (examples; hydrochloric, hydrobromic,
hydriodic, iodic, nitric, and perchloric acids, primary ionisation of sulphuric
acid), and the latter are called weak acids (examples: nitrous acid, acetic acid,
carbonic acid, boric acid, phosphorous acid, phosphoric acid, hydrocyanic acid,
and hydrogen sulphide). There is, however, no sharp division between the two
classes.

A base may be defined as a substance which, when dissolved in water,
undergoes dissociation with the formation of hydroxide ions OH~. as the only
negative ions. Thus sodium hydroxide, potassium hydroxide, and the hydroxides
of certain bivalent metals are almost completely dissociated in aqueous solution :

NaOH-^Na+ 4 -OH-
Ba(0H)2^Ba+ + -t- 20 H-

These are strong bases. Aqueous ammonia solution, however, is a weak base. Only
a small concentration of hydroxide ions is produced in aqueous solution :

NH3 + H2O ^ NH4 + + OH -

General concept of acid and bases. The Brnnsted theory. The simple
Arrhenius concept given in the preceding paragraphs suffices for many of the
requirements of quantitative inorganic analysis in aqueous solution. It is,
however, desirable to have some knowledge of the general theory of acids and
bases proposed by J. N. Bronsted in 1923 , since this is applicable to all solvents.
According to Bronsted, an acid is a species having a tendency to lose a proton,
and a base is a species having a tendency to add on a proton. This may be
represented as:

Acid Proton + Conjugate Base
A^H+-t-B

(«)

It must be emphasised that the symbol (p'*" is sometimes used) represents the

proton and not the ‘hydrogen ion’ of variable nature existing in different solvents
> NH4 , CH3C02H2^, C2H50H2^, ctc.); the definition is therefore
independent of solvent. The above equation represents a hypothetical scheme for
defining A and B and not a reaction which can actually occur. Acids need not be
neutral molecules (e.g., HCl, H2SO4, CH3CO2H), but may also be anions (e.g
p ’ HOOC COO ) and cations (e.g., NH4+, C6H5NH3+!

r eiHzUjg ). The same is true of bases where the three classes can be illustrated

by NH 3 , C6H3NH2, H2O; CH3COO-, OH-, HPO. 2 - OCH-

Fe(H20)5(0H)^+. 3 , , nrij4 , UU2H5 ,

rr. 1 QUANTITATIVE INORGANIC ANALYSIS

Since ilic free proton cannot exist in solution in mcasiirr-.hif.
reaction docs not take place unless a base is , ‘^‘’"“"'’’ation,

acid. By combining the equations • * • ccpi the proton from the

A,c*B,+ir and

weobtam Aj+b, ^A, + n, ' ' ''

*" » 1 ...

Aj-B] and Ai-B^ arc tsvo coniucaie acid -In*:.' n iirc "n • • t

expression for reactions involvinu acids and ‘ ‘™PO'''ant

proton from A, to B, or from A 1 ^ 1 ^ Thl? ^ 'he transfer ofa

A,, the more complete will be the reaeVion U,) Ti"'^T ‘ ’’'c weaker

more readily than the weaker- similarly the st’ml sut'nger acid loses its proton
readily than does the weaker base It is evident ihafo a proton more

strong acid or a strong base is always xveak wher‘-a! ’f 'o ‘‘

a weak acid or weak base is alwavs'strone ‘ conjugate to

In aqueous solution a Bronsicd acid A^ '

A+H,0=iH,0* 4-R

A s^bSonS’::;^;;!!^

Tcfds m.?’'" hydroxide ion, is almoJ.So concentration of

tendencies \Wth a base, which for aoiieon^^r'^f'"" **^'^*'^ rchitive combining
'ntcrcstcd) is water: ‘'tqiitous solutions (in which we arc largely

Hci-fH,o=iH,oN- cr

CH.COOH + H,0 a H ,o . 1 - * " “I™'''’” ^

aL.

+cn3Coo

B.-iic,

prod„c/„“/,;!;?e, ~ » «;.si ttT ”'T 'pf"''™

" * js the ionic

-repair may involve

pxtended clasVof - in!? “'j ^ principle ^ HjO- OH ■). showing

important plaee in nrn''f'-'^ bases, though, of course the '-’■^‘‘B’ples of an

he greatly innueneed h ^°hows that the nronn r ‘ P-’riicularlv

16 ’ "bet on indicators, catalysis).

FUNDAMENTAL THEORETICAL PRINCIPLES II, 2

irrespective of their chemical nature and mode of action, as acids or bases. He
relates the properties of acids to the acceptance of electron pairs, and bases as
donors of electron pairs, to form covalent bonds regardless of whether protons
are involved. On the experimental side Lewis’ definition brings together a wide
range of qualitative phenomena, e.g., solutions of BF3, BCI3, AICI3, or SO 2 in an
inert solvent cause colour changes in indicators similar to those produced by
hydrochloric acid, and these changes are reversed by bases so that titrations can
be carried out. Compounds of the type of BF 3 are usually described as Lewis acids
or electron acceptors. The Lewis bases (e.g., ammonia, pyridine) are virtually
identical with the Bronsted bases. The great disadvantage of the Lewis definition
of acids is that, unlike proton-transfer reactions, it is incapable of general
quantitative treatment.

Salts. The structure of numerous salts in the solid state has been
investigated by means of X-rays and by other methods, and it has been shown
that they are composed of charged atoms or groups of atoms held together in a
crystal lattice. When these salts are dissolved in a solvent of high dielectric
constant such as water, or are heated to the melting point, the crystal forces are
weakened and the substances dissociate into the pre-existing charged particles or
ions, so that the resultant liquids are good conductors of electricity. There are,
however, some exceptions: feebly ionised salts (weak electrolytes) are exemplified
by the cyanides, thiocyanates, and halides of mercury and cadmium, and by lead
acetate.

The theoretical implications of the theory of complete ionisation, due to
Debye, Huckel, and Onsager, have been fully worked out by these authors. In
particular, they have been able to account for the increasing equivalent
conductance with decreasing concentration over the concentration range 0-
0.002M. For full details the reader must be referred to textbooks of physical
chemistry.

It is important to realise that whilst complete ionisation occurs with strong
electrolytes, this does not mean that the effective concentrations of the ions are the
same at all concentrations, for if this were the case, the osmotic properties of
aqueous solutions could not be accounted for. The variation of osmotic
properties with dilution is ascribed to changes in the activity of the ions; these are
dependent upon the electrical forces between the ions. Expressions for the
variations of the activity or of related quantities, applicable to dilute solutions,
have also been deduced by the Debye-Hiickel-Onsager theory. Further
consideration of the concept of activity will be found in Section II, 3.

n, 2. THE LAW OF MASS ACTION. Guldberg and Waage in 1 867 clearly
stated the law of mass action (sometimes termed the law of chemical equilibrium)
m the form: the velocity of a chemical reaction is proportional to the product of
the active masses of the reacting substances. For the present ,we shall interpret
active mass’ by concentration and express it in mols per cubic decimetre. By
applying the law to homogeneous systems, i.e., to systems in which all the reacting
molecules are present in one phase, for example in solution, we can arrive at a
mathematical expression for the condition of equilibrium in a reversible reaction.

Let us consider first the simple reversible reaction at constant temperature:

A-|-B:?^C-|-D

The velocity with which A and B react is proportional to their concentrations, or

n, 3 QUANTITATIVE INORGANIC ANALYSIS

t'l [A] X [B]

where k, is a constant known as the velocity coefficient, and the square brackets
(sec p, 30 footnote) denote the molecular concentrations of the substances
enclosed within the brackets. Similarly, the velocity with which the reverse
reaction occurs is given by;

At equilibrium, the vdocitic.s of the reverse and the forward reactions will be
equal (the equilibrium is a dynamic and not a static one) and therefore r, - Cj,

or /;,x{A]xlR!-kjX[Clx(D!

[C]x[D ] _ k,
lAlxlBJ " k,

K is the equilibrium constant of the reaction at the given temperature.

The expression may I>c generalised, l-'or a reversible reaction represented by:

+Pi'^> Tr/jB; + ‘/jBj -t-... .

where Pi, pj, pj and (jj, r/j, are the number of rnoleculc.s of reacting substances,
the condition for equilibrium is given by the expression :

[Br'x[B;I'- x[B,r-.... ^ ^

{A,fx(Ajpx[A,K‘....

This result may be expressed in words : when cq uilibrium is reached in a reversible
reaction, at constant temperature, the product of the molecular concentrations of
the resultants (the substances on the right-hand side of the equation) divided by
the product of the molecular concentrations of the reactants (the substances on
the left-hand side of the equation), each concentration being raised to a power
equal to the number of molecules of that substance taking part in the reaction, is
constant.

11, 3. ACTIVnV AND ACTIVITY COEFITCIENT. In our deduction of
the law of mass action it svas assumed that the effective concentrations or active
masses of the components could he cxprc.sscd bv the stoichiometric con-
cptrations. According to modern thermodynamics, this is not strictlv true. The
rigorous equilibrium equation for, say, a binary electrolyte:

AB5±:A*-fB-

(oa- Xfln-)

-fin

where^.,a„.,and ci,,„ represent llic aetivilic,s of A \ B ' , and AB respectively, and
IS the true or thermodynamic dis.sociation constant. The concept of activity, a
quantity, is due to G. N. Lewis. The quantity is related to the
concentration by a factor, termed the activity coefficient :

activity = concentration x activity coefficient
Thus at any concentration

Oa- =>'a*.(A*], <7b- =yn-.[B''],andfl^n= Vab

.{AB]

FUNDAMENTAL THEORETICAL PRINCIPLES D, 4

where y refers to the activity coefficients,* and the square brackets to the
concentrations. Substituting in the above equation, we obtain :

yA-.[A+]xyB- [B,“] _ [A'*']. [B"] y** x T b- _
yAB-[AB] [AB] , Tab “

This is the rigorously correct expression for the law of mass action as applied to
weak electrolytes.

The activity coefficient varies with the concentration. For ions it also varies
with the valency, and is the same for all dilute solutions having the same ionic
strength, the latter being a measure of the electrical field existing in the solution.
The term ionic strength, designated by the symbol I, is defined as equal to one half
of the sum of the products of the concentration of each ion multiplied by the
square of its valency, or / = 0.5ZC(Z,.^, where c,- is the ionic concentration in mols
per cubic decimetre of solution and z,- is the valency of the ion concerned. An
example will make this clear. The ionic strength of O.IM-HNO 3 solution
containing 0 . 2 M-Ba(NO 3)2 is given by:

0.5{0.1 (forH+)+0.1 (forN 03 -)

+ 0.2 X 2^ (for Ba^ + ) + (0.2 x 2) (for NOj")} = 0.5{ 1 .4} = 0.7.

The activity coefficient depends upon the total ionic strength of the solution in
a manner which is discussed in Section II, 8. The activity coefficients of un-ionised
molecules do not differ considerably from unity and for weak electrolytes in
which the ionic concentration and therefore the ionic strength is small, the error
introduced by neglecting the difference between the actual values of the activity
coefficient of the ions, and yg-, and unity is small ( < 5 per cent). Hence for
weak electrolytes, the true or thermodynamic expression reduces to [A'^] x [B~]/
[AB] = K, and the constants obtained by the use of simple concentrations will be
accurate to 2-5 per cent; such values are sufficiently precise for many of the
calculations related to quantitative analysis.

n, 4 . ACID-BASE EQUILIBRIA IN WATER. Let us consider the
dissociation of a weak electrolyte, such as acetic acid, in dilute aqueous solution :

CH3COOH + H2O ^ H3O + + CH3COO “

This will be written for simplicity in the conventional manner:

CH 3 C 00 H^H+ +CH3COO"

where H represents the hydrated hydrogen ion. Applying the law of mass action,
we have:

[CH3COO-] X [H+]/[CH3C00H] = JC

K is the equilibrium constant at a particular temperature and is usually known as
the ionisation constant or dissociation constant. If one mol of the electrolyte is

The symbol used is dependent upon the method of expressing the concentration of the solution. The
recommendations of the lUPAC Commission on Symbols, Terminology and Units (1969) are as
follows: concentration in mols per cubic decimetre (molarity), activity coefficient represented by y,
concentration in mols per kilogram (molality), activity coefficient represented by y, concentration
expressed as mole fraction, activity coefficient represented by/.

FUNDAMENTAL THEORETICAL PRINCIPLES II, 5

and the constant

[A][H, 0 ]

( 2 )

gives the strength of A, that of the.ion being taken as unity. Equation (b)

represents what is usually described as the dissociation of the acid A in water, and
the constant K' is closely related to the dissociation constant of A in water as
usually defined and differing only in the inclusion of the term [H2O] in the
denominator. The latter term represents the ‘concentration’ of water molecules in
liquid water ( 55.5 raols per cubic decimetre on the ordinary volume concentration
scale). When dealing with dilute solutions, the value of [H2O] may be regarded as
constant, and equation ( 2 ) may be expressed as:

'■'TaT

( 3 )

by writing H'*' for and remembering that the hydrated proton is meant.

This equation defines the strength of the acid A. If A is an uncharged molecule
(e.g., a weak organic acid), B is the anion derived from it by the loss of a proton,
and ( 3 ) is the usual expression for the ionisation constant. If A is an anion such as
H2P04’,the dissociation constant [HP04^"][H'*']/[H2P04~] is usually referred
to as the second dissociation constant of phosphoric acid. If A is a cation acid, for
example the ammonium ion, which interacts with water as shown by the equation

NH4+ + H20^NH3 + H 30 +

the acid strength is given by [NH3][H'^]/[NH4''].

On the above basis it is, in principle, unnecessary to treat the strength of bases
separately from acids, since any protolytic reaction involving an acid must also
involve its conjugate base. The basic properties of ammonia and various amines
in water are readily understood on the Bronsted concept.

H20 = H++0H-

NH3+H+ =NH4 +

nhT+h^o^nh^^ToiF

The basic dissociation constant K,, is given by:

[NH 4 +][ 0 H-]

[NH3]

( 4 )

where [NH3] represents the total concentration of ammonia, irrespective of
whether it is present as free NH3 or as NH4OH; no reliable evidence is available
as to the actual existence of NH4OH. Since [H+][OH"] = K,, (the ionic product
ofwater),wehave

Kh = KJK,

The values of K„ and Kf, for different acids and bases vary through many
de^^^d convenient to use the dissociation constant exponent pK

P^ = logiol/J*C= -logioX

the larger the pK„ value the weaker the acid and the stronger the base.

II, 6 QUANTITATIVE INORGANIC ANALYSIS

For very weak or slightly ionised electrolytes, the expression aVd -a)V =
reduces to a* = KFor a = .^/KV, since a may be neglected in comparison with
unity. Hence for any two weak adds or bases at a gjvcn^lution F(in dm^), we
havc«i = == Expressed in words,

for any two weak or slightly dissociated electrolytes at cc]Uiil dilutions, the
degrees of dissociation arc proportional to the square roots of their ionisation
constants. Some values for the dissociation constant.s at 25 “C for weak adds and
bases are collected in Appendix XI.

11,6. DISSOCIATION OF POLYPROT IC (POLYIIASIC) ACIDS. When

a polyprotic add is dissolved in water, the various hydrogen atoms undergo
ionisa’tion todifferent extents. For a diprotic add H ^ A, the primary and secondary
dissociations can be represented by the equations;

HjA^iH" +HA" («)

HA-^HN-A^' (h)

If the acid is a weak electrolyte, the law of mass action may be applied, and the
following expressions obtained:

[H-']x[A‘')/lHA-l= (2)

fv, and K 2 arc known as the primary and secondary dissociation constants

respectively. Each stage of the dissociation process has its own ionisation
constant, and the magnitudes of these constants give a measure of the extent to
which each ionisation has proceeded at any given concentration. The greater the
value of Kj relative to Xj, the smaller will be the secondary dissociation, and the
greater must be the dilution before the latter becomes appreciable. It is thereorc
possible that a diprotic (or polyprotic) acid may behave, so far as dissodnlion is
concerned, as a monoproticacid. This is indeed characteristic of many polyprotic
acids.

A triprotic acid HjA (e.g., orthophosphoric acid) will similarly yield three
dissociation constants, f.',. K^.and Xj, which may be computed in an analogous
manner:

HjA^iir + HjA- (c)

HjA- +HA^- (J)

\c)

We can now apply some of the theoretical considerations to actual examples
encountered in analysis.

ExampleX. ^oC'tIc'tl''^cthcconccntraiionsofHS” andS’" inasolulionof
hydrogen sulphide.

A saturated aqueous solution of hydrogen sulphide at 25 'C, at atmospheric
pressure IS approximately O.IM, and for 11,S the primary and sccondary
d^sociation constants may be taken as l.Ox 10'^ moldm'^ and 1 x lO"*" mol
am respectively.

In the solution the following equilibria are involved :

(c)

{/)

ig)

FUNDAMENTAL THEORETICAL PRINCIPLES II, 7

Electroneutrality requires that the total cation concentration must equal total

anion- concentration and hence, taking account of valencies,'

[Hi = [HS-]+2[S2"]+[OH“] {h)

but since in fact we are dealing with an acid solution, [H’^] > > [OH“]and

we can simplify equation (h) to read

lH+] = [HS-] + 2[S2"] (j)

The 0.1 mol H 2 S is present partly as undissociated H^S and partly as the ions
HS~ and S^", and it follows that

[H2S] + [HS-] + [S2-] = 0.1 W

The very small value of K 2 indicates that the secondary dissociation and
therefore [S^“] is extremely minute, and ignoring [S^ “j in equation (/) we are left
with the result

[H+] « [HS-] ' ' , , ’ , (0

Since is also small, [HjS] and so equation (k) can be reduced to

[HjSJ^O.l ' N

Using these results in the expression for we find

[H+p/0.1 = 1x10-’; [H+] = [HS-]= 1.0x10-“' mol dm-^ -

From equation (/) it then follows that

(1.0 X 10-^)[S’-]/(1.0 X 10-^) = 1 X 10-
and [S^-] = lxlO '“'mol dm

IfwemultiplyKibyX2wefind[S2-] = 1 x10-2V[H+]'^.

Thus the concentration of the sulphide ion is inversely proportional tothe square
of the hydrogen-ion concentration, i.e., if we, say, double [H ''■] by the addition of a
strong acid, the [S’ -] will be reduced to or j of its original value.

n, 7, COMMON ION EFFECT. The concentration of a particular ion in an
ionic reaction can be increased by the addition of a compound which produces
that ion upon dissociation. The particular ion is thus derived from the compound
already in solution and also from the added reagent, hence the name common ion.
We shall confine our attention to the case in which the original compound is a
weak electrolyte in order that the law of mass action may be applicable. The result
IS usually that there is a higher concentration of this ion in solution than that
derived from the original compound alone, and new equilibrium conditions will
be produced. Examples of the calculation of the common ion effect are given'
belo w. In general, it may be stated that if the total concentration of the common
ion is only slightly greater than that which the original compound alone, would
furnish, the effect is small; if, however, the concentration of the common ion is
very much increased (e.g., by the addition of a completely dissociated salt), the
effect is very great, and may be of considerable practical importance. Indeed, the
common ion effect provides a valuable method for controlling the coricentration
of the ions furnished by a weak electrolyte. . . ,

Example 2. To calculate the sulphide-ion concentration in a 0.25M-
hydrochloric acid solution saturated with hydrogen sulphide.

11, 7 QUANTITATIVK INORGANIC ANALYSIS

This concentration has been chosen since it is that at which tlic sulphides of
certain heavy metals arc precipitated. The total concentration of hydrogen
sulphide may be assumed to be approximately the same as in aqueous solution,
i.c., O.IA/: tlt’e [IP] will be equal to that of the completely dissociated HCI, i.e.,
Q.25M, but the [S^ “] will be reduced below 1 v 10 " ’■*.

Substituting in equations («) and {b) (Section H. 6 ). we find :

L',x!H,S) 1.0x10'- xO.l ,,

IHS'1 = --- — ~ — ~ 4.0 X 10 mo! dm •

‘ I [u>|

0.25

x-U

[S^-

[H^j

0.25

Thus by changing the acidity from 1,0 x 10'‘‘Af (that present in saturated HjS
water) to 0,25Af. the sulphide-ion concentration is reduced from I x 10 ' to 1.6
xlO-^'.

Example ?>. What efl'ect has the addition of 0,1 mol of anhydrous sodium
acetate to 1 cim^ of 0.1 Af-acclic acid upon the degree of dissociation of the acid?

Tlic dissociation constant of acetic acid at 25 C is 1.S2 x lO"'' mo! dm" ^ and
the degree of ionisation a in 0 . 1 , Af solution may be computed by solving the
quadratic equation:

[HCjHjOT]

a^f

(T^i)

1.82 .X 10 “

For our purpose it is sulhcicntly accurate to neglect a in (1 —a) since a is small:
a = ^/kTc = = 0.0135

Hence in 0 . 1 M -acetic acid,

[H^] = 0.00135, [CjHjO;"] = 0.00135. and [H-CjHjO,) - 0.0986 mol dm"^

The concentrations of sodium and acetate ions produced by the addition of the
completely dissociated sodium acetate are;

[Na — 0.1, and (CjUjO, '] = 0.1 mol dm " ^ respectively.

The acetate ions from the salt will tend to decrease the ionisation of the acetic
acid, and consequently tlie acetate-ion concentration derived from it. Hence we
may write [C 3 H 3 O 2 ] = 0.1 for the solution, and if ct' is the new degree of
ionisation, [H ’'] = a'c = 0 . 1 a', and (H-CjHjO,] = (1 ~a)c - 0 . 1 , since a' is
negligibly small.

Substituting in the mass-action equation:

[brjxlCjHjOj-] 0.1a' X 0.1

(H-CjHjO,] 0l “1-S2xl0'5

or a'= 1.8x10"*

[H'*^] — a'e = 1,8 x ! 0 ~* mol dm"^.

The addition of a tenth of a mole of sodium acetate to a 0. 1 M solution of acetic
acid has decreased the degree of ionisation from 1.35 to 0,018 per cent, and the
hydrogen-ion concentration from 0.00135 to 0.000018 mol dm"-\

Example 4. What effect has the addition of 0.5 mol of ammonium chloride

FUNDAMENTAL THEORETICAL PRINCIPLES H, 8

to 1 dm^ of O.lM-aqueous ammonia solution upon the degree of dissociation of
the base?

(Dissociation constant ofNHj in water = 1.8x10“^ mol dm"^)

In . O.lM-ammonia solution a = y^l.8 x 10~V0-1 == 0.0135.. Hence
[OH"] = 0.00135, [NH4'^] = 0.00135, and [NH3] = 0.0986 mol dm"^. Let a' be
the degree of ionisation in the presence of the added ammonium chloride. Then
[OH"] = a'c = 0.1a', and [NH3] = (1.— a')c = 0.1, since a' may be taken as
negligibly small. The addition of the completely ionised ammonium chloride will,
of necessity, decrease the [NH4'''] derived from the base and increase [NH3], and
as a first approximation [NH4‘^] = 0.5.

Substituting in the equation :

[NH4 "]x[OH-] _ 0.5 X 0.1a'

[NH3] 0.1

a' = 3.6 X 10“^ and [OH~] = 3.6 x 10“® mol dm

The addition of half a mole of ammonium chloride to a O.IM solution of
aqueous ammonia has decreased the degree of ionisation from 1.35 to 0.0036 per
cent, and the hydroxide-ion concentration from 0.001 35 to 0.0000036 mol dm "

n, 8. SOLUBILITY PRODUCT. For sparingly soluble salts (i.e., those of
which the solubility is less than 0.01 mol per dm^) it is an experimental fact that
the product of the total molecular concentrations of the ions is a constant at
constant temperature. This product Ks is termed the solubility product For a
binary electrolyte;

AB^A+ + B“

A;,ab) = [A+] X [B-]

In general, for an electrolyte ApB^, which ionises into pA’’*' and qB'’~ ions;
ApB^^pA’+Fq-B""

Vb.)» = [A«^]'’x[Bp-]«

A plausible deduction of the solubility product relation is the following. When
excess of a sparingly soluble electrolyte, say silver chloride, is shaken' up with
water, some of it passes into solution to form a saturated solution of the salt and
the reaction appears to cease. The following equilibrium is actually present (the
silver chloride is completely ionised in solution) ;

AgCl (solid) :^Ag'^-t-Cr

The velocity of the forward reaction depends only upon the temperature and at
any given temperature;

H = ■

where is a constant. The velocity of the reverse reaction, is proportional to the
activity of each of the reactants ; hence at any given temperature ;

= kj X X Uc-

where fcj is another constant. At equilibrium the two velocities are equal, i.e.

n, 8 QUANTITATIVE INORGANIC ANALYSIS

/ij ~ ^2 X X iIq-
or X (i^ = ~

In the very dilute solutions with whicli we arc concerned, the activities may be
taken as practically equal to the concentrations so that lAg ' ] x (Cl " ) ^ const.

It is important to note that the solubility product relation applies with
suflicient accuracy for purposes of quantitative analysis only to saturated
solutions of slightly soluble electrolytes and with snwH additiorrs of other salts. In
the presence of inodcnite concentrations of salts, the ionic concentration, and
therefore the ionic strength of the solution, will increase. Thi.s will, in general,
lower the activity cod'iicicnts of both ions, and consequently the ionic
concentrations {and therefore the solubility) must incretisc in order to maintain
the solubility product constant. Tliiscflcct, which is ntosl marked w hen the added
electrolyte docs not possess an ion in common with the sparingly soluble salt, we
may term the salt elTcct. It can be .shown on the basis of the Debye-Huckcl-
Onsager theory that for aqueous solutions at 25 ^C;

logy,=

where y, is the activity coeflicicnt of the ion, r, is the valency of the ion concerned, /
is the ionic strength of the solution (.Section IL 3), and a is the average ‘clTectisc
diameter’ of all the ions in the solution. For very dilute soluiion.s (/^'^ < 0.1) the
second term of the denominator is negligible and the equation reduces to:

logy, = -0.505r,*./"'^

For more concentrated solutions (f” ^ > 0.3) an additional term B/ is added to
the equation; B is an empirical constant. For a more detailed treatment of tlic
influence of salts upon solubility and solubility product, the reader is referred to
textbooks of clcctroclicmistry.

It will be clear from the above short discussion that two factors may come into
play w'hcn a solution of a salt containing a common ion i.s added to a saturated
solution of a slightly soluble salt. At moderate conccntnitions of the added salt,
the solubility will generally decrease, but with higher concentrations of the
soluble salt, when the ionic strength of the solution increases considerably and
the activity coefficients of the ions decrease, the solubility may actually increase.
This is one of the rca.sons why a very large c.xccss of the precipitating agent is
avoided in quantitative analysis.

A few examples may help the reader to fully understand the subject. The
concentrations are expressed in mols per dm* for the calculation of solubility
products.

Example 5. Tlic solubility of silver chloride is 0.001 S g per dm*. Calculate
the solubility product.

n weight of silver chloride is 143.5. The solubility is therefore

0.0015/143.5 = I.05x 10 mol per dm*. In a saturated solution, 1 nioIeofAgCl
wil give 1 mole each o Ag^ and CP. Hence [Ag-] = 1.05 x 10"* and [CP]
= 1.05 X 10 * mol dm'*.

= 1.1 xlO'*® moF

FUNDAMENTAL THEORETICAL PRINCIPLES O, 9

Example 6. Calculate the solubility product of silver chromate, given that
its solubility is 2.5 X 10" ^ g per dm^

Ag2Cr04^2Ag++Cr04-

The molecular weight of Ag 2 Cr 04 is 332, hence the solubility = 2.5 x 10 “ 7332
= 7.5 X 10“^ mol dm “

Now 1 mole of Ag 2 Cr 04 gives 2 moles of Ag"*^ and 1 mole of Cr 04 ^";
therefore

WK>.) = [Ag"]^x[Cr04^]

= (2x7.5 X 10" 7^ X (7.5x10"
= 1.7 X 10"’^ moP dm“7

Example 7. The solubility product of magnesium hydroxide is 3.4 x 10 “
moP dm"®. Calculate its solubility in grams per dm^.

Mg(0H)2^Mg2 + + 2OH"

[Mg"+]x[OH“]^ = 3.4xlO-*‘

The molecular weight of magnesium hydroxide is 58. Each mole of magnesium
hydroxide, when dissolved, yields 1 mole of magnesium ions and 2 moles of
hydroxyl ions. If the solubility is x mol dm"^, [Mg^'''] = x and [OH"] = 2x.
Substituting these values in the solubility product expression :

XX (2x)^ = 3.4 X 10"^^
or X = 2.0 X 10"''^mol dm"^

= 2.0xl0"^x58 = 1.2x10"^ gdm"^

The great importance of the solubility product concept lies in its bearing upon
precipitation from solution, which is, of course, one of the principal operations of
quantitative analysis. The solubility product is the ultimate value which is
attained by the ionic concentration product when equilibrium has been
established between the solid phase of a difficultly soluble salt and the solution. If
the experimental conditions are such that the ionic concentration product is
different from the solubility product, then the system will attempt to adjust itself
m such a manner that the ionic and solubility products are equal in value. Thus, if,
for a given electrolyte, the product of the concentrations of the ions in solution is
arbitrarily made to exceed the solubility product, as, for example, by the addition
of a salt with a common ion, the adjustment of the system to equilibrium results in
precipitation of the solid salt, provided supersaturation conditions are excluded.
If the ionic concentration product is less than the solubility product or can
arbitrarily be made so, as, for example, by complex salt formation or by the
ormation of weak electrolytes, then a further quantity , of solute can pass into
solution until the solubility product is attained, or, if this is not possible, until all
the solute has dissolved.

n. 9. QUANTITATIVE EFFECTS OF A COMMON ION. An important
solubility product principle is to the calculation of the
so ubility of sparingly soluble salts in solutions of salts with a common ion. Thus
e solubility of a salt MA in the presence of a relatively large amount of the

II, 9 QUANTITATIVE INORGANIC ANALYSIS

common M"" ions,* supplied by a second sail MB. follows from the definition of
solubility products:

lM*]x|A-J=

or IA'I-AWTM] H)

Tlie solubility of the salt is represented by the (A “j which it furnishes in solution.
It is clear that the addition of a common ion willdemvMv thcsolubility of the salt.

Esample S. Calculate the .solubility of silver chloride in (o) 0.001 M- and (/;)
O.OlAf-sodium chloride solutions respectively

In a saturated solution of silver chloride (CP ) ~ \.l x lO" = 1.05 x 10"-'

mol dm"-': this may be nenlccted in comparison with the excess of Cl" ions
added.

Forf«)[Cr] - 1 xlO*'.jAg'l - 1.1 :< iO"'"/! >: 10'^

— 1.1 X 10"" mol dm"-'

ForlhllCl'] = 1 X lO'pfAg") l.l r i0-'“/l x 10" =

= l.l X lO'-’* mot dm" •'

Thus the solubility is decreased 100 limes in 0.001 A/-sodium chloride and 1000
times in 0.01 Af-sodium chloride. Similar results are obtained for O.tXllAf- and
0.01 Af -silver nitrate solution.

Examph' 9. Calculate the solubilities of silver chromate in 0.001 A/- and
0.01 Af- silver nitrate solutions, and in 0.001 A/-and O.OI .\/-poinssium chromate
solutions (Ag2Cr04: A', = 1.7 x 10" '= mob' dm'“, solubility in water =
7.5 X 10'® mol dm"-').

(Ag']=x(Cr04=')= 1.7xl0-‘ =
or (Cr04="l = 1.7x10" ‘=4Ag‘]=

For 0.001 Af-silver nitrate solution ; [Ag ■‘I = ! x 10" '

ICrO*^"] = 1,7 X 10" '=/! X 10'*' = 1.7 x 10' *'moI dm”'.
For 0,01 Af-silver nitrate solution ;(Ag‘] = 1 x 10" =

[Cr04= ] — . 1.7 X 10 ’*/! X 10 ■* = 1,7 lO'^mol dm"'.
The solubility product equation gives:

[Ag*] = /l.7 X 10" '=/fCr04=’" [

For [Cr04=-] - 0.001, (Ag*) = ^/T? xlO" ■=/! ^iF"'
= 4.1 X 10 moidm"'.

For [Cr04 = "] = 0.01, [Ag* ] =

= 1.3 X 10"' mol dm"'.

This decrease in solubility by the common-ion effect is of fundamental
importance in gravimetric analysis. By the addition of a suitable excess of a
precipi ^ titg the solubility of a precipitate is usually decreased to so small
a value that the loss from solubility innucnccs is negligible. Let us consider a
specific case— the determination of silver as silver chloride. Here the chloride

This enables us to neglect the conccmralion ofM*
.and Ihus to simplify the calculation.

ions .supplied by the .sparingly .solubles, ill itself.

FUNDAMENTAL THEORETICAL PRINCIPLES II, 10

solutionis added to the solution of the silver salt. If an exactly equivalent amount
is added, the resultant saturated solution of silver chloride will contain 0.0015 g
per dm^ {Example 5). If 0.2 g of silver chloride is produced and the volume of the
solution and washings is 500 cm^, the loss, owing to solubility, will be 0.00075 g or
0.38 per cent of the weight of the salt ; the analysis would then be 0.38 per cent too
low. By using an excess of the precipitant, say, to a concentration of 0.01 M, the
solubility ofthe silver chloride is reduced to 1.5 x 10"^ gperdm^(£xamp/e8),and
the loss will be 1.5 x 10~® x 0.5 x 100/0.2 = 0.0038 per, cent. Silver chloride is
therefore very suitable for the quantitative determination of silver with high
accuracy.

It should, however, be noted that as the concentration of the excess of
precipitant increases, so too does the ionic strength of the solution. This leads to a
decrease in activity coefficient values with the result that to maintain the value of
more of the precipitate will dissolve. In other words there is a limit to the
amount of precipitant which can be safely added in excess. Also, addition of
excess precipitant may sometimes result in the formation of soluble complexes
causing some precipitate to dissolve.

n, 10. FRACTIONAL PRECIPITATION. We have thus far considered the
solubility product principle in connection with the precipitation of one sparingly
soluble salt. We shall now extend our studies to the case where two slightly
soluble salts may be formed. For simplicity, we shall study the situation which
arises when a precipitating agent is added to a solution containing two anions,
both of which form slightly soluble salts with the same cation, e.g., when silver
nitrate solution is added to a solution containing both chloride and iodide ions.
The questions which arise are; which salt will be precipitated first, and how
completely will the first salt be precipitated before the second ion begins to react
with the reagent?

The solubility products of silver chloride and silver iodide are respectively 1.2
X lO'^^moF dm“® and 1.7 x 10”^® moP dm“®;i.e.,

[Ag+]x[Cr] = 1 . 2 x 10 “^° ( 1 )

[Ag+]x[I-] = 1.7x10-^® (2)

It is evident that silver iodide, being less soluble, will be precipitated first since its
solubility product will be first exceeded. Silver chloride will be precipitated when
the Ag'*' ion concentration is greater than

^s(AgCl) 1.2xl0~^°

[cr] [cP]

and then both salts will be precipitated simultaneously. When silver chloride
commences to precipitate, silver ions will be in equilibrium with both salts, and
equations (1) and (2) will be simultaneously satisfied, or

and

[Ag+1 = — .
[I-]

P ] _ j^gi)

■^s(AgCl)

■ [cr]

1.7x10-

= 1.4x10"®

(3)

(4)

[Cl ] i^s(Aga) 1.2 X 10 • ' '

Hence when the concentration of the iodide ion is about one millionth part of the
c londe-ion concentration, silver chloride will be precipitated. If the initial

II, 11 QUANTITATIVE INORGANIC ANALYSIS

concentration of both chloride and iodide ions is 0. 1 M, then silver chloride wi] be
precipitated when

[r] = 0.1 X 1 . 4 x 10 -'^’ = 1.4x I0"\M = l.Sx lO'^gdm"^

Thus an almost complete separation is theoretically possible. Tlic separation is
feasible in practice if the point at wliieh the iodide precipitation is complete can be
detected. This may be done : (a) by tbc u.sc of an adsorption indicator (see Section
X, 30C), or (b) by a poicntiomctric method with a silver electrode (sec Ch.apter
XIV).

For a mixture of bromide and iodide;

[Br-j 3.5 xio-’^ 2.0x10'

Precipitation of silver bromide will occur when the concentration of the bromide
ion in the solution is 2.0 x 10’ times that of the iodide concentration. Tlie
separation is therefore not quite so complete as in the ca.se of chloride and iodide,
but can ncvcrthclc-ss be cficctcd with fair accuracy with the aid of adsorption
indicators (Section X, 30C).

11,11. COMPLF.X IONS. The increase in solubility of a predpitate upon the
addition of excess of the precipitating agent is frequently due to the formation of a
complex ion. A complex ion is formed by the union of a .simple ion with cither
other ions of opposite charge or with neutral molecules. Let us examine a few
examples in detail.

When potassium cyanide solution is added to a solution of silver nitrate, a
white precipitate of silver cyanide is first fonned because the solubility product of
silver cyanide:

{Agn^(CN-] = AWN, (1)

is exceeded. The reaction is expressed:

CN'+Ag" =AgCN

The precipitate dissolves upon the addition of excess of potassium cyanide, the
complex ion [AglCN),)" being produced;

AgCN (solid) + CN' (excess) =±(Ag(CNK] ' «

(orAgCN + K-CN = K.[Ag(CN).) — a soluble complex salt)

This complex ion dissociates to give silver ions, since the addition of sulphide ions
yields a precipitate of silver .sulphide (solubility product 1.6 .x 10 mol’ dm'"),
an a so silver is deposited from the complex cyanide solution upon clectrolvsis.
The complex ion thus dissociates in accordance wiiii the equation;

[AgfCN)^]- :^Ag^ +2CN"

fhTwhok purposes : to denote concentrations and also to include

careful scrulinv there sh”" i '"''r tirackeis (braces) arc sometimes used. With

used:withco,nple.xesth=re.iU^°nri;aSS^

FUNDAMENTAL THEORETICAL PRINCIPLES H, 11

Applying the law, of mass action, we obtain the dissociation constant of the
complexion:

' ■’ [Ag1x[CN-f _ ^ ■ ■ ' ,21

[{Ag(CN),}-]

which has a value of l.Ox moP dm~® at the ordinary temperature. By
inspection of this expression, and bearing in mind that excess of cyanide ion is
present, it is.evident that the silver ion concentration must be very small, so small
in fact that the solubility product of silver cyanide is pot exceeded.

The inverse of equation (2) gives us the stability constant or formation constant
of the complex ion: ...

[{Ag(CN)r}]

[Ag+]x[CN-f

= 10^^ mol ^ dm®

( 3 )

Consider now a somewhat different type of complex -ion formation, viz., the
production of a complex ion with constituents other than the common ion present
in the solution. This is exemplified by the solubility of silver chloride in ammonia
solution. The reaction is :

AgCl-l-2NH3 ^[Ag(NH3)2]+ -pCr

Here again, electrolysis, or treatment with hydrogen sulphide, shows that silver
ions are present in solution. The dissociation of the complex ion is represented by:

[Ag(NH3)2]+^Ag+-f-2NH3
and the dissociation constant is given by:

_ [Ag+]x[NH3]^

[{Ag(NH3)2}"]

= 6.8 X 10 ® moFdm ®.

The stability constant K = = 1.5 x 10’ mol ^dm®.

The magnitude of the dissociation constant clearly shows that only a very small
silver ion concentration is produced by the dissociation of the complex ion.

The stability of complex ions varies within very wide limits. It is quantitatively
expressed by means of the stability constant. The more stable the complex, the
greater is the stability constant, i.e., the smaller is the tendency of the complex ion
to dissociate into its constituent ions. When the complex ion is very stable, e.g.,
the hexacyanoferrate(II) ion [Fe(CN)6]‘* ~> Ihs ordinary ionic reactions of the
components are not shown.

The application of complex-ion formation in chemical separations depends
upon the fact that one component may be transformed into a complex ion which
IS no longer precipitable with the precipitating agent, whereas another
component is precipitated. On'b example may be mentioned here. This is
concerned with the separation of cadmium and copper. Excess of potassium
cyanide solution is added to the solution containing the two salts when the
complex ions [CdfCN)^]^" and [Cu(CN) 4 ]^~ respectively are formed. Upon
passing hydrogen sulphide into the solution containing excess of CN“ ions, a
precipitate of cadmium sulphide is produced. Despite the higher solubility
product of CdS (1.4 x 10“^® mol’ dm“® as against 6.5 x 10“'^® mol’ dm~® for
copper sulphide), the former is precipitated because the complex cyanocuprate(I)

rr, 12 QUANTITATIVE INORGANIC ANALYSIS

stabiliiy constant (2y. mol"-*

xJO mol-* dm‘ = for the cadmium compound) compared with 7

(Secr„sI 9 a 7 "'“'“°''°'"

siBSsSS™'

M"+A-+H^-HA + M'

If the dissociation constant of the -iHH H s ■-

removed from the solution to form ih- r ^mall, the anion A* will h-

Td tlds‘'’“'"' I’'''‘'^ ">'csoLtion to'rSShf-r'

Mrochlorie acid is if t ^A) or ’i ^ ffidL

soluble salt has dissS

acid A, = 7.5x 10-^ mol dm*^' A' ?n l'""'^'‘‘-''''*'^uchasp

™°>dm ^),oxalicacid(A'. =^9 J A% -= <;xT-u

Tc acid, Thus ilJcni.i:,.^.'^, nioldm-^:A', =. 6,4 x lO-^moldm-^j,

„ mnasHPO,- and/or H,PO,-- :

V +H z±uvn z- • ‘‘

HPO^^' + ir^lilp^C^

,I.andmtrous(A'. =J6y in-4 , <^'01 ■\- A', = J Ox 10“ '

me In !i,o ; ', ^-Os^IO mol dm ■ •ii'i/ir

• IS the arm..! !u '*^'^*''onal factor

contributina o ■ ■ ^

'■romsolutioneii fr "^^'ubiliiy is the acin-.i' T‘""’ "" '‘"‘'•''cnai lactor

provided fof f w5r;;r--'>- or Jr. gerlde t -‘id

provided for the vvum "'^“"’’'^ rirongcnticuarmin
carbonates oxahi ■‘’dubilitv of "^^■A"'^'^P>‘'nationisthus

“»P.ionV

PIfAg(CN),l. 'diich ■

*1," i'-"-'. ij-ide;' ,3"

S04‘'4-H"-hso “ ^-^iO-inoIdm-^):

Since, however AT 1 ^: r

precipitate and to re^duee solution in order to Precipitation mav

Section XI, lo. s^rbstances within a contrnllVi’

controlled range of pH is discussed-

FUNDAMENTAL THEORETICAL PRINCIPLES H; 13/14/15

n, 13. EFFECT OF TEMPERATURE UPON THE SOLUBILITY OF A
PRECIPITATE. The solubility of the precipitates encountered in quantitative
analysis increases with rise of temperature. With some substances the influence of
temperature is small, but with others it is quite appreciable. Thus the solubility of
silver chloride at 10 and 100 °Cis 1.72 and 21.1 mg dm“^ respectively, whilst that
of barium sulphate at these two temperatures is 2.2 and 3.9 mg dm " ^ respectively.
In many instances, the common-ion effect reduces the solubility to so small a
value that the temperatures effect, which is otherwise appreciable, becomes very
small. Wherever possible it is advantageous to filter while the solution is hot; the
rate of filtration is increased, as is also the solubility of foreign substances, thus
rendering their removal from the precipitate more complete. The double
phosphates of ammonium with magnesium, maganese or zinc, as well as lead
sulphate and silver chloride, are usually filtered at the laboratory temperature to
avoid solubility losses.

n, 14. EFFECT OF THE SOLVENT UPON THE SOLUBILITY OF A
PRECIPITATE. The solubility of most inorganic compounds is reduced by the
addition of organic solvents, such as methanol, ethanol, and propan-l-ol, acetone,
etc. For example, the addition of about 20 per cent by volume of ethanol renders
the solubility of lead sulphate practically negligible, thus permitting quantitative
separation. Similarly calcium sulphate separates quantitatively from 50 per cent
ethanol. Other examples of the influence of solvent will be found in Chapter XI.

n, 15, THE IONIC PRODUCT OF WATER. Kohlrausch and Heydweiller
(1894) found that the most highly purified water that can be obtained possesses a
small but definite conductivity. Water must therefore be slightly ionised in
accordance with the equation :

H20^H++0H-*

Applying the law of mass action to this equation, we obtain, for any given
temperature;

Ah* ^ Aqh-

[H^].[OH ] yH* .yoH-
[HjO] yH,o

= a constant

Since water is only slightly ionised, the ionic concentrations will be small, and
their activity coefficients may be regarded as unity; the activity of the unionised
molecules may also be taken as unity. The expression thus becomes:

[H+]x[OH-]

= a constant

1^2 UJ

In pure water or in dilute aqueous solutions, the concentration of the
undissociated water may be considered constant. Hence:

Strictly speaking the hydrogen ion exists in water as the hydroxonium ion (Section II, 1).
e electrolytic dissociation of water should therefore be written :

2Hj0^H30''+0H-

For the sake of simplicity, the more familiar symbol H'*' will be retained.

II, 16 QUANTITATIVE INORGANIC ANALYSIS

where A'„, is the ionic prodticl of water. It must be pointed out that the assumption
that the activity cocflicicnts ofihe ions are unity and that the activity cocnidcnl of
water is constant applies strictly to pure water and to very dilute solution.s (ionic
strength c 0.01); in more concentrated solutions, i.e., in solutions of appreciable
ionic strength, the electrical environment afTcct.s the activity coeflicients of the
ions (compare Section If, 8) and also the activity of the un-ioni.scd water. The
ionic product of water will then not be con.siant, but will depend upon the ionic
environment. It is, however, dinicult to determine the activity coelTicicnts, c.xccpt
under specially selected conditions, so that in practice the ionic product K.^,
although not strictly constant, is employed.

The ionic product varies with the temperature, but under ordinary'
c.xpcrimental conditionsfat about 25 ‘ Qit.s value may be taken a.s 1 x 10* '■* with
concentrations expressed in mol This is sensibly constant in dilute

aqueous solutions. If the product of [H‘j and {OH*) in aqueous solution
momentarily exceeds this value, the excess ions will immediately combine to form
water. Similarly, if (he product of the two ionic concentrations is momentarily less
than 10“ more water molecules will dissociate until the equilibrium value is
attained.

The hydrogen- and hydroxide-ion concentrations arc equal in pure water;
therefore - [OH*] ~ ~ 10' ’ mol dm*’ at about 25 'C. A solution

in which the hydrogen- and hydroxide-ion concentrations are equal is termed an
exactly neutral solution. If (H * J is greater than 10 * the solution i.s acid, and if less
than 10" \ the solution is alkaline (or lia.srci. It follows that at ordinary
temperatures [Oil '] is greater than 10* ' in alkaline .solution and less than this
value in acid solution.

In ail eases the reaction of the solution can be quantitatively expa'ssed by the
magnitude of the hydrogen-ion (or hydroxonium-ion) concentration, or. less
frequently, of the hydroxide-ion concentration, since the following simple,
relations between [H j and [OH ' j cxi.st ;

The variation of with temperature is shown in Table II. 1.

Table II, I. Ionic Product of XVnter at \ arioas Tempera lures

Tcrap.(’C)

A'.xlO'*

Tcni(i.( C)

A. X J0'‘

10'

15”

20’

25’

30’

0.12

0.19

0.29

0.45

0.68

1.01

1.47

35'

40’

45’

50'

55-

2.09

2.92

4.02

5,47

7.30

9,61

n, 16. THE HYDROGEN-ION FVPnMrvT r-

neristle wbor. LAFONENT. For many pun>o.se.s, es-

con^lralions of hydioj™™” 'bSrf'-""'' ■' “

dad.e,.. A c„L.r„" “J s':s„s.“oi

FUNDAMENTAL THEORETICAL PRINCIPLES II, 16

He introduced the hydrogen-ion exponent pH defined by the relationships:

pH = logic I/[H+] = -logioPn, or [H+] = IQ-o”

The quantity pH is thus the logarithm (to the base 10) of the reciprocal of the
hydrogen-ion concentration, or is equal to the logarithm of the hydrogen-ion
concentration with negative sign. This method has the advantage that all states of
acidity and alkalinity between those of solutions eontaining on the one hand, 1
mol dm“^ of hydrogen ions, and on the other hand, 1 mol dm"^ of hydroxide
ions, can be expressed by a series of positive numbers between 0 and 14. Thus a
neutral solution with [H"^] = 10“’ has a pH of 7; a solution with a hydrogen ion
concentration of 1 mol dm“^ has a pH of 0 ([H^] = 10°); and a solution with a
hydroxideionconcentrationof 1 moldm“^has[H'''] = lt„,/[OH"] = 10“*V10°
= 10“ and possesses a pH of 14. A neutral solution is therefore one in which
pH = 7, an acid solution one in which pH < 7, and an alkaline solution one in
which pH > 7. An alternative definition for a neutral solution, applicable to all
temperatures, is one in which the hydrogen-ion and hydroxide-ion concen-
trations are equal. In an acid solution the hydrogen-ion concentration exceeds
the hydroxide-ion concentration, whilst in an alkaline or basic solution, the
hydroxide-ion concentration is greater.

Example 10. (i) Find the pH of a solution in which [H'*^] = 4.0 x 10“ ^ mol
dm“^.

pH = logio l/[H+3 = logl -log[H+]

= log 1.— log 4.0 X 10 ^

= 0-5.602
= 4.398

(ii) Find the hydrogen ion concentration corresponding to pH = 5.643.

pH = logio 1/[H+] = log 1 -log[H+] = 5.643
.-. log [H+]= -5.643

This must be written in the usual form containing a negative characteristic and a
positive mantissa:

log[H+]= -5.643 = 6.357

By reference to tables of antilogarithms we find [H ''' ] = 2.28 x 10~° mol dm “

(iii) Calculate the pH of a 0.90 IM solution of acetic acid in which the degree of
dissociation is 12.6 per cent. The hydrogen ion concentration of the solution is
0.125x0.01

= 1.25 X 10“° mol dm“°.
pH = logio 1/[H+] = log 1 -log[H+]

= 0-3.097
= 2.903 .

The hydroxide-ion concentration may be expressed in a similar way:
pOH= -logio[OH“] = logio l/[OH“], or [OH-]= 10“P°» : ’

If we write the equation:

[H+]x[OH“] = Ji:„= 10“!'^

Ill 17 QUANTlTATIVn JNORGAN’rC ANALYSIS

in llie form;

'os(H'l + Iog[Off'J = IoPA' ^ -14

then pH + pOH = pf:.,J ,4

™ir n Mrilil"'™.''’ n, nbou, 25 .r

|OH-i,a„d pon il, acWandilt!,tas"SoS°" "" I*' '!• P".

T' 'f f‘ ‘ f f«- R"

I ■•>.■67*0 in ,1 ,,

pon 14 13 12 „ ,,, 9 j i , , , , I

FIs- n, I

AfValn^

si''??'™ “■'‘•'■"I f"

P^a = l0gl/A',= -log^:__

Similar,, r„ra„,- ba» d'^r^aapoa

P ^ ~ tog I/A'^ = _log/j^

lii) ''“'••>")ion;orco„c«„„„„„„p,.

P'-i»eMi|= -login
^"'•'"'Na'I-S^lo-.pNa^g,

P “lobimy,.rod„c, A- .

P'^.-logI/K,=

main dasSf "™''0*-VSIS OF SALTS,

1- 'hose derived from ri ■ ' ' ^ '"'o four

3- Ih'Se' S;S flom T'' Po'assinm chloride;

< .^e derived Lo " """

™.™t‘“™fS'ir' “—»<»"■ ronnaic or

r SrsS" -i. hnown. is no,

sail.

FUNDAMENTAL THEORETICAL PRINCIPLES H, 17

With an aqueous solution of a salt of class (1), neither the anions have any
tendency to combine with the hydrogen ions nor the cations with the hydroxide
ions of water, since the related acids and bases are strong electrolytes. The
equilibrium between the hydrogen and hydroxide ions in water:

HjO^H+tOH- («)

is therefore not disturbed and the solution remains neutral.

Consider, however, a salt MA derived from a weak acid HA and a strong base
BOH {class (2)}. The salt is completely dissociated in aqueous solution :

MA — >M++A“

A very small concentration of hydrogen and hydroxide ions, originating from the
small but finite ionisation of water, will be initially present. HA is a weak acid,
that is, it is dissociated only to a small degree; the concentration of A“ ions which
can exist in equilibrium with H"^ ions is accordingly small. In order to maintain
the equilibrium, the large initial concentration of A ' ions must be reduced by
combination with H"^ ions to form undissociated HA:

H++A-^HA (b)

The hydrogen ions required for this reaction can be obtained only from the
further dissociation of the water; this dissociation produces simultaneously an
equivalent quantity of hydroxyl ions. The hydrogen ions are utilised in the
formation of HA, consequently the hydroxide ion concentration of the solution
will increase and the solution will react alkaline. The net result is that the anions
of the salt react with the hydrogen ions of the water, yielding the weak acid HA,
and there is an increase in the concentration of hydroxide ions over that present
in water.

It is usual in writing equations involving equilibria between completely
dissociated and slightly dissociated or sparingly soluble substances to employ the
ions of the former and the molecules of the latter. The reaction is thefore written :

A- + H20:^0H- + HA (c)

This equation can also be obtained by combining (a) and (b), since both equilibria
must co-exist. This interaction between the ion (or ions) of a salt and the ions of
water is called hydrolysis.

Let us now study the salt of a strong acid and a weak base (class (3)} . Here the
initial high concentration of cations M''" will be reduced by combination with
the hydroxide ions of water to form the little dissociated base MOH until
the equilibrium :

M+ + 0H-^M0H (d)

IS attained. The hydrogen-ion concentration of the solution will thus be
increased, and the solution will react acid. The hydrolysis is here represented by:

M+-fH20^MOH-|-H+ (e)

For salts of class (4), in which both the acid and the base are weak, two re-
actions will occur simultaneously

M+-t-H20:?±M0H-f-H+ (f)

A--fH20-HA + OH- /„

II. 18 QUANTITATIVE INORGANIC ANALYSIS

The readion of ihc soHnion will clearly depend upon ihc relative dissociation
constants of the acid and the base. If they arc equal in strength, thcsohition will be
neutral; if > Kf,. it will be acid, and if > K,, it will be alkaline.

Having considered all the possible cases, we are now in a position to give a
more nenera! definition of hydrolysis. Hydrolysis is the interaction between an
ion (or ion.s) of a salt and the ion.s of water with the production offo) a wetik acid
or a weak base, or {h) of both a weak acid and a weak ba.se.

The phenomenon of sail hydrolysis may be regarded as a simple application of
the general Bronsicd equation

A, + B; + B)

Thus the equation for the hydroly.sis of ammonium .s.ilis

NH." 4-Hj0e:iNH,, + H,0‘

is really identical with tlie expression used to deline the strength of the
ammonium ion as a Bronsled acid (see Section II. 5) and the con.siani for
NH 4 * is in fact what is usually termed the hydrolysis constant of an ammonium
salt.

The hydrolysis of the sodium salt of a weak acid c;in he treated similarly. Tims
for a solution of sodium acetate

CHjCOO'+HjO^iCHjCOOH + OlI .
the hydrolysis constant is

(CnjCOOHKOH'l.'lCHjCOO ) - K, KjK,
where is the dissociation con.stant of acetic acid.

ir, 18. HYDROU’SIS CONSTANT AND DKGREK OF HYDRO-
LYSIS. Case 1 . Salt of a Weak acid and a strong base.

Tlie equilibrium in a solution of a salt M A may be represented bv:

A - +H,Oe:±OH +HA

Applying the law of mass action, sve obtain:

^ .'bn-.Viu (1)

“A- [A

where is the hydroly.sis constant. The solution is assumed to be dilute so that
the activity of the unioni.scd water may be taken as constant, and the
approximation that the activity cocflkient of tlie un-ionised acid is unity and that

both ions have the same activity cocfiicient may be introduced. Equation (1) then
reduces to: ' -i ' '

,, [OH-)>:(l-IA1

(Pj (2)

This is often written in the form:

j. ^ [Base] X [Acid]

[OnhydrdP^^dSaiT]

the free strong base and tlie unhydroU'sed
acid is very little dissociated.

salt arc completely dissociated and the

FUNDAMENTAL THEORETICAL PRINCIPLES II, 18

The degree of hydrolysis is the fraction of each mole hydrolysed at equilibrium.
Let 1 mol of salt be dissolved in V dm^ of solution, and let x be the degree of
hydrolysis. The concentrations in mols dm " ^ are :

A' +H20;^0H"+HA

(i-xyv x/v xjv

Substituting these values in (2);

[OH"]xIHA] x/Fxx/F x^

. [A-] - (l-x)/F "(l-x)F' , , .

This expression enables us to calculate the degree of hydrolysis at the dilution F;
it is evident that as F increases, the degree of hydrolysis x must increase;

The two equilibria:

H20^H+ + 0H-
and HA^H++A-

must co-exist with the hydrolytic equilibrium;

A-+H20^HA-t-OH-

Hence the two relationships;

IH+]x[OH-] = K,„
and [H+]x[A-]/[HA] = K,

must hold in the same solution as:

[OH-]x[HA]/[A-] = JC,

Rnt ^ ^ PA] [OH-] X [HA] ^

K, [H+]x[A-] [A-] "

therefore KJK, =

or pK, = pK,„-pK„

The hydrolysis constant is thus related to the ionic product of the water and the
ionisation constant of the acid. Since K„ varies slightly and varies
considerably with temperature, and consequently the degree of hydrolysis will
be largely influenced by changes of temperature.

The hydrogen-ion concentration of a solution of a hydrolysed salt can be
readily computed. The amounts of HA and of OH“ ions formed as a result of
hydrolysis are equal, therefore in a solution of the pure salt in water [HA]
= [OH ]. If the concentration of the salt is c mol dm then:

[^]x[OH-] [OH-f
[A-] c

and [0¥l-] = ^c.KJK^ ■ ■ . ■

[H+] = Jk^.KJc, since [H+] = KJ[OH-]
and pH = ipi;:„+ipi;:^+ilogc

To be consistent we should use pc = — logc. .

FUNDAMENTAL THEORETICAL PRINCIPLES II, 18

[H'^]x[MOH]_ [Acid]x[Base] _

~ [M^ [Unhydrolysed Salt] Kj,

Xj,is the dissociation constant of the base. Furthermore, since [MOH] and [H"^]
are equal;

_ [H^]x[MOH] [H-^f K,,

'• [M+] c K,

[H+] = ^c.KJK,,

or pH = ipK„,-ipKj.+ipc (4)

Equation (4) may be applied to the calculation of the pH of solutions of salts of
strong acids and weak bases. Thus the pH of a solution of ammonium chloride
(0.2 mol dm~^) is;

pH = 7.0-2.37 (0.70) = 4.98

(Ammonia in water: — 1.85 x 10~^ mol dm~^; pKj = 4.74)

Case 3. Salt of a weak acid and a weak base.

The hydrolytic equilibrium is expressed by the equation :

M^+A' + HjO^MOH + HA

Applying the law of mass action and taking the activity of un-ionised water as
unity, we have:

2^ . ^MOH ^ %A _ [MOH] . [HA] Tmoh'Tha
[M+].[A"] ^ yM*-yA-

By the usual approximations, that is, by assuming that the activity coefficients of
the un-ionised molecules and, less justifiably, of the ions are unity, the following
approximate equation is obtained ;

^ _[MOH]x[HA]

* 1m+]x[A-] ■

_ [Base] X [Acid]

[Unhydrolysed Salt]^

fix is the degree of hydrolysis of 1 mol of the salt dissolved in V dm^ of solution,
then the individual concentrations are :

[MOH] = [HA] = x/F; [M+] = [A”] = (1 -x)/r
leading to the result

xjV.xlV x"-

- (1

11, 18 QUANTITATIVE INORGANIC ANALYSIS

pH =

Equation (3) can be employed for the calculation of the pH of a solution of a
salt of a weak acid and a strong base. Thus the pH of a solution of sodium
benzoate (0.05 mol dm“^) is given by;

pH = 7.0+2.10-^(1.30) = S.45

(Benzoic acid: = 6.37 x 1 0” ^ mol dm ~ ^ ~ 4.20)

Such a calculation will provide useful information as to the indiaitor which
should be employed in the titration of a weak acid and a strong base (see Section
X, 13).

Example 11. Calculate: (i) the hydrolysis constant, (ii) the degree of
hydrolysis, and (iii) the hydrogen-ion concentration of a solution of .sodium
acetate (0.01 mol dm“'^)at the laboratory temperature.

h —

1.0x10"'*

=: 5.5 X 10'"’

The degree of hydrolysis .v is given by;
^ (1-a-)H

Substituting for and I’ { = l/c), we obtain:

(1-.V)

Solving this quadratic equation for x, .v = 0.000235 or 0.0235 per cent.
CjHjOj'-f HjOeiH CjHjOj-f OH'

(l—.x) moles .t moles .tmolfs

If the solution were completely hydrolysed, the concentration of acetic acid
produced would bcO.Oi mol dm " But ihedcgree ofhydrolysis is 0.0235 percent,
therefore the concentration of acetic acid is 2.35 >; 10"*' m'ol dm~^. Tliis is also
equal to the hydroxide-ion concentration produced, i.e., pOH = 5.63.

pH = 14.0-5.63 = 8.37

The pH may also be computed from equation (3):

pH ==lpX„-flpX,-Jpc

= 7.0 + 2.37-4(2) = 8.37.

Case 2. Salt of a strong acid and a weak base.

The hydrolytic equilibrium is represented by:

M-’+HjOqiMOH + H*

By applying the law of mass action along the lines of Case 1, the following
equations are obtained :

FUNDAMENTAL THEORETICAL PRINCIPLES II, 18

[H'*'] X [MOH] _ [Acid] x [Base] _

~ [Unhydrolysed Salt] K^,

'{l-x)V

Ki is the dissociation constant of the base. Furthermore, since [MOH] and [H”^]
are equal:

[H+]x[MOH] ,

[M+] c K,

[H+] = Jc.KJK,,

or pH = - ipiCfc + ipc (4)

Equation (4) may be applied to the calculation of the pH of solutions of salts of
strong acids and weak bases. Thus the pH of a solution of ammonium chloride
(0.2mol dm~^)is;

pH = 7.0-2.37+1(0.70) = 4.98

(Ammonia in water: Kj = 1.85 x 10~^ mol dm"^; pR:^ = 4.74)

Case 3. Salt of a weak acid and a weak base.

The hydrolytic equilibrium is expressed by the equation ;

M^ + A'+HjO^^MOH + HA

Applying the law of mass action and taking the activity of un-ionised water as
unity, we have:

_ °MOH X %A ^ [MOH] . [HA] yMOH'THA
Om^xoa- [M+].[A-] yM*-TA-

[MOH] X [HA]

[M+]x[A-]

_ [Base] X [Acid]

[Unhydrolysed Salt]^

If X is the degree of hydrolysis of 1 mol of the salt dissolved in V dm^ of solution,
then the individual concentrations are;

[MOH] = [HA] = x/F; [M+] = ja'] = (i _x)/K
leading to the result

II, 19 quantitative inorganic analysis

The degree of hydrolysis and consequently the pM is indcr)enf!,.nf r t.
concentration of the solution.* * 'Ricpcndcnt of the

It may be readily shown that:
or pA's = pA\.-pA'„-pAV

tli^issociation constant's of th "S and^h‘c^ ofhydrolysis .x from

/olloIwSSn " concentration of the hydrdy.sed solution is calculated in the

(H^] = A'„x

(HA]

= A'

y -

x/f

rr = A',

(i-x)/p-

But.v/(l-x) = /A';

hence firicu-A' /F - /ir — r--:;-
or PH=ipA\.fipA',_i,p;^.^

ir (he ionisaiion consianr^ nf »h » • ■ i

A'a = A\,pH = ^ 7 0 ‘•od the ba.se arc equal, that is.

be considerable. If A > A nil "^‘“"’‘‘’•‘■''‘’’ough hvdrolysis may

Tl., pH ofa ^

PH = 7.0 + 2..17-2.37 = 7.0

of ammonium fonnaic:"'"'"''"^''’ ^’"hcr hand, for a dilute solution

pH = 7.0 + 1 . 8 S_ 2.37^0,,

1,7,

> e., the solution reacts Slightly acid. ^ “

n, 19. BUFFER SOLinriONS \ , •

CCS of alkali from the glass of the con ""'"“oo 't: extremely sensitive to

J>ave a pH ono^p"'"" ''vdro;rfo'S' ;r! T

Aqueous soh,. ’ 'o traces of carh,-n. 7 mol dm' ^), which should

about 7 The chloride and of the atmosphere.

bydrochlonV ' dm-’ of tlti

and in vp.-,, i dm"-') rcsuli.s in „ J of a solution of

bufforsoin ,,S,^'°'"'‘°"''’^'ich

solutions usuailvrn Ve.servc acidfiv' I Rcopcrtics is known as a

usually con.sist ofsolutions containin ' ^ uikalinity’. Buffer

r- ^ "‘‘‘'"'"P uuxture of a weak acid HA and

Applies Only if thp .

PPrecablc .on,c strengtl, ,hc ■•>«iviiy are ju.s.ificd. In ..olniions of

42 «ry with Iho total ionic strength.

FUNDAMENTAL THEORETICAL PRINCIPLES II, 19

its sodium or potassium salt (A ~), or of a weak base B and its salt A buffer,

then, is usually a mixture of an acid and its eonjugate base. In order to understand
buffer action, let us study first the equilibrium between a weak acid and its salt.
The dissociation of a weak acid is given by :

HA^H++A-

and its magnitude is controlled by the value of the dissociation constant K„:
an* ^

^H* —

%A ^A-

( 1 )

The expression may be approximated by writing concentrations for activities

1H*1 = PJxK. (2)

This equilibrium applies to a mixture of an acid HA and its salt, say MA. If the
concentration of the acid be and that of the salt be c^, then the concentration of
the undissociated portion of the acid is c^ — The solution is electrically
neutral, hence [A“j = Cs + [H'^] (the salt is completely dissociated). Substituting
these values in the equilibrium equation (2), we have;

[HI

c.+[Hn

(3)

This is a quadratic equation for [H'*'] and may be solved in the usual manner. It
can, however, be simplified by introducing the following further approximations.
In a mixture of a weak acid and its salt, the dissociation of the acid is repressed by
the common ion effect, and [H"^] may be taken as negligibly small by comparison
with Ca and c^. Equation (3) then reduces to:

[H+]

[Acid]

[Salt]

(4)

The equations can be readily expressed in a somewhat more general form when
applied to a Bronsted acid A and its conjugate base B:

A^H+ + B

(e g., CH 3 COOH and CHjCOO”, etc.). The expression for pH is:

pH = pK„ + log|^
where K, = [H+][B]/[A].

Similarly for a mixture of a weak base of dissociation constant and its salt

with a strong acid:

[OH-]

[Base]

lSalt]‘

xK,

pOH = pK,+iog

[Salt]

[Base]

( 6 )

(7)

II, 19 QUANTITATIVI- INORGANIC ANALYSIS

Let us confine our attention to the ctisc in which the concentrations of the acid
and its salt are equal, i.c., of a half-neutralised acid. Then pi F ~pK^. Thus the pH
of a half-neutralised .solution of a weak acid is equal to the negative logarithm of
the dissociation constant of the acid. For acetic acid, “ 1.82 x 10 ■ mol
dm" \ piva - 4.74; a half-neutralised solution of, say. 0. 1 Af-acctic acid will have a
pH of 4.74. If we add a small conc-cniration of H'* ions to such a solution, the
former will combine with the acetate tons to form undissociated acetic acid;

H^-bCHaCOO" ^CHjCOOH

Similarly, if a small concentration of hydroxide ions be added, the latter will
combine wath the hydrogen ions arising from the dissociation of the acetic acid
and form un-ionised water; the equilibrium will be disturbed, and more acetic
acid will di.ssociate to replace the hydrogen ions removed in this way. In either
ease, the concentration of the acetic acid and acetate ion (or .salt) will not be
appreciably changed. It follows from equation (5) that the pM of the solution will
not be materially affected.

Example 12. Calculate the pH of the solution produced by adding 10 cm^
of 1 M hydrochloric acid to 1 dm’ of a solution which is O.I Af in acetic acid and
0.1 Af in sodium acetate = 1. 82 10'’ mol dm ' ').

The pH of the acetic acid-sodium acetate buffer solution i.s given by the
equation;

pH = pf^a + !ogp---r,l = 4.74 + 0.0 -• 4.74
[Acid]

The hydrogen ions from the hydrochloric acid react with acetate ions forming
practically undissociated acetic acid, and neglecting the change in volume from
1000cm’ to lOlOcm’ weean say

CHjCOO' =^0.1 -O.OI =0.09
CHjCOOH =0,1+0.01 =0,11

and pH = .t .74 h- log 0.09/0.1 1 .- 4.7-1 -0.09 = 4.65.

Thus the pH of the acetic acid sodium acetate buffer solution is only altered by
0.09 pH unit on the addition of the hydrochloric acid, llie same volume of
hydrochloric acid added to one litre of water (pH = 7) would lead to a .solution
With pH - -log(0.01) = 2; a change of 5 pH unit.s. This example serves to
illustrate the regulation of pH c.\erci.sed bv buffer solutions.

A solution containing equal concentrations of acid and its salt, or a half-
neu raliscd solution of the acid, has the maximum buffer capacity. Other
mixtures also po.sscss considerable bu(fcr capacity, but the pH will differ slichtly

add [Acidfi 3 [Salti"''“^^ 'l"-'‘'‘‘^r-ncutr.aliscd solution of

pH = pK„ + logi =pfv'^+].52

= pA:„-0.48

Fora threc-quartcr-ncutraliscd acid, (Salt] = 3 (Acid);

pH = pA:„ + log3
= pA'„-+0.48

In general, we may state that the buffering capacity is maintained for mixtures

FUNDAMENTAL THEORETICAL PRINCIPLES 11, 19

within the range 1 acid: 10 salt and 10 acid : 1 salt. The approximate pH range of a
weak acid buffer is:

pH = pK,±l

The concentration of the acid is usually of the order 0.05-0.2 mol dm“‘’. Similar
remarks apply to weak bases.

The preparation of a buffer solution of a definite pH is a simple process if the
acid (or base) of appropriate dissociation constant is found: small variations in
pH are obtained by variations in the ratio of the acid to the salt concentration.
One example is given in Table II, 2.

Table n, 2. pH of Acetic Acid-Sodium Acetate Buffer Mixtures
lOcm^mixluvesofxcm^ ofO.lM-aceticacid and )>cm* ofO.lM-sodium acclatc

Acetic Acid (x cm*) Sodium Acetate (y cm*) pH

9.5 0 5 3.42

9.0 1 0 3.72

8.0 2 0 4.05

7.0 3.0 4.27

6.0 4.0 4.45

5.0 50 4.63

4.0 6 0 4.80

3.0 7^0 4.99

2.0 8.0 5.23

1.0 9.0 5.57

0.5 9.5 5.89

n. 20 QUANTITATIVE INORGANIC ANALYSIS

where r.- is (lie valency of the ion, I is the ionic strength of the solution, and A and
B are constants. This may be written in the form

logy, = -0.505-'.^/“-' -i-C/

where C is another constant approximately equal to 0.?05c,*.d Tfl and usually
has a value varying between 0.2 anti 1.5. Substituting for y^- in (10), we obtain:

pH = pA', + log [Saltj.'lAcid]- 0.505c, + (11)

Tlie activity cocfTicient of the ion \\- generally increases with decrease of
concentration, so that when a buffer solution is diluted, Va increases and
consequently will increase {equation (9)}. Tor mo.st practical purposes the
change in pH is small, but for exact work it must be taken into account. The
addition of salts to buffer mi.xttires results in a change of the ionic strength of the
solution; this will affect the pH of the soluiton {equation (11)). Indeed, in all
buffer solutions a correction shouhi, strictly .speaking, lie applied for the ionic
strength oflhe solution.

Buffer mixtures arc not confined to mixtiirc,s of monoprotic acid.s or monoacid
bases and their salts. We may employ a mixture of .salts of a polyprotic acid, e.g,,
NaHjPOi and NajUPO*. The salt NaHjPOA is completely di.ssociatcd:

NaHjPO^eiNa' 4 -H,POa"

The ion HjPOa* acts as n monoprotic acid;

HjPOa' eirir -) HPOa‘-

for which K{ s for phosphoric acid) is 6.2 x 10 ' mol dm ' The addition of
the salt Na,HP 04 is analogous to the addition of, say, acetate ions ton solution
of acetic acid, since the tertiary ionisation of phosphoric acid {HPOi*“ ;iH^
+ P 04 ^“) is .small (K, - 5 x lb' mol dm"^). The mixture of NaHsPO^ and
NajIiPO^ is therefore an clfcclive buffer over the range pH 7.2± 1.0(^pK' ± I).
It will be noted that this is a mixture of a Bronsted acid and its conjugate basc.

Buffer solutions find many applications in quantitative inorganic analysis, e.g.,
many precipitations are only quantitative under carefully controlled conditions
of pH, as are also many complcximetric titrations; numerous cxnmplc,s of their
use will be found throughout the book.

I, 20. ELECTRODE rorENTfAI.S. When a metal is immersed in a
soKition containing its own ion.s, say. zinc m /.me sulphate solution, a potential
dmercncc is established between the metal and the solution. The potential
difiercnce £ for an electrode reaction

is given by the expression:

( 1 )

legasconstant, Tistheabsolutc temperature. /'thcFaradayconstant,
constanf*^'!tTn^ H ^ ‘''^livity of the ions in the solution, and is a

introdiirino Equation (1) can be simplified by
g le nown values of R and F. and converting natural logarithms to

FUNDAMENTAL THEORETICAL PRINCIPLES H, 20

base lOby multiplying by 2.3026; it then becomes:
0.0001983 r'

- log Am-

For a temperature of 25 °C (T = 298K):

_ 0.0591,'

E = E^-{ logflM"*

( 2 )

For most purposes in quantitative analysis, it is sufficiently accurate to replace
Am., by Cm-*, the ion concentration (in moles per dm^):

_ 0.0591 ,

E = E^+ log Cm-

( 3 )

The latter is a form of the Nernst equation.

If in equation (2), is put equal to unity, E is equal to £®. is called the
standard electrode potential of the metal.

In order to determine the potential difference between an electrode and a
solution, it is necessary to have another electrode and solution of accurately

known potential difference. The two electrodes can
then be combined to form a voltaie cell, the e.m.f. of
which can be directly measured. The e.m.f. of the
cell is the arithmetical sum or difference of the
electrode potentials (depending upon the sign of
these two potentials); the value of the unknown
potential can then be calculated. The primary
reference electrode is the normal or standard hy-
drogen electrode (see also Section XTV, 2). This
consists of a piece of platinum foil, coated electro-
lytically with platinum black, and immersed in a
solution of hydrochloric , acid containing hydrogen
ions at unit activity. (This corresponds to 1.8M-
hydrochloric acid at 25 °C.) Hydrogen gas at a
pressure of one atmosphere is passed over, the
platinum foil through the side tube C (Fig. II, 2) and
escapes through the small holes B in the surrounding glass tube A. Because of the
periodic formation of bubbles, the level of the liquid inside the tube fluctuates,
and a part of the foil is alternately exposed to the solution and to hydrogen. The
ower end of the foil is continuously immersed in the solution to avoid
interruption of the electric current. Connection between the platinum foil and an
external circuit is made with mercury in D. The platinum black has the
property of adsorbing large quantities of hydrogen, and it permits
e change from the gaseous to the ionic form and the reverse process to occur
without hindrance; it therefore behaves as though it were composed entirely of
y rogen, that is, as a hydrogen electrode. Under fixed cbnditions, viz;,
y rogen gas at atmospheric pressure and unit activity of hydrogen ions in the
so ution in contact with the electrode, the hydrogen electrode possesses a definite
Po ential. By convention, the potential of the standard hydrogen electrode is
temperatures. Upon connecting the standard hydrogen
e ec rode with a metal electrode (a metal in contact with a solution of its ions of

Hg-II,!

II, 20 QUANTITATIVE INORGANIC ANALYSIS

unit activity) by means of a salt (say, potassium cltloride) bridge, the standard
electrode potential may be determined, the cell i.s usually written as

Pt.Hjiirio- 1)11M"M«= DjM

In this scheme, a single vertical line represents a mclal-clcciroiytc boundary at
which a potential dilTercnce is taken into account; a double vertical line
represents a liquid junction at which the potential is to be disregarded or is
considered to be eliminated by a salt bridge.

When we speak of the electrode potential of a ?inc electrode, we mean the e.m.f.
ofthcccll;

Pt.H;iHMa= l)i|Zn=‘ jZn
or the e.m.f. of the half-cell Zn’ * j Zn, The cell reaction is:

llj + Zn=* — .2I-r (« = D-t-Zn
and the half-cell reaction is written as:

Zn^ " -t-2e:;:iZn

The electrode potential of the Fe^". Fc* j Pt electrode is the e.m.f. of the cell:

PLHjjHMo= l)j|Fc^*,Fc*'iPt
or the e.m.f of the half-cell Fc’ ’ ,Fc’ * [ Pt. The cell reaction is;

’Hj + Fc^" — 1-r ( 0 = n-t-Fc-’’
and the half-cell reaction is written;

Fc^-^ + c'=iFc=*

The convention is tidoptcd of writing all half-cell reactions ns reductions;

M'"'

c.g., 7.n'*+2i':^ Zn ip _ _ o.76 volt

When the activity of the ion M"’ is equal to unity (appro.ximntelv true for a \M
solution), the electrode potential E is equal to the sttindard potential Some
important standard electrode potentials referred to the standard hydrogen
electrode at 25 "C (in aqueous solution) are collected in Table 11. 3.

Table 11, 3. Standard Electrode I’olcnticls ai 2S C

Electrode reaction

E‘^ (voliq

Li* +c = Li

-3.045

K*-rc = K

-2.925

Ba^ * +2e = Ba

-2.90

Sr^* +2f =rSr

-2.89

Ca=-* +2c = C.T

-2.87

Rt* -(-e = Nn

-2.714

Mg-’*+2c = Mr*

-2.37

= A!

-1,66

Mn^* -t-2e ^ Mn

- 1. IS

Zn^*+2c = Zn

-0.763

Ec^* +2e = Fc

-0.440

Cd^*+2(. = Cd

-0403

nwtrede rraciion

E^t^otts)

T1 * Xl

-0.336

Co*’H-2or7Co

-0,277

-0.25

Sn’’ + 2«* ■=; Sn

-0.136

* '4'3c =s Pb

-0.126

2H;

0000

Cw* * -V 2f «. Cu

-1 0.337

« Up

4-0,7K9

Ap^ *f c jf* Ag

+ 0.799

-i 2c ^ Pd

+ 0.9S7

*= Pt

+3c e= Au

+ 1.2
+ 1.50

FUNDAMENTAL THEORETICAL PRINCIPLES H, 21/22

It may be noted that the standard hydrogen electrode is rather difficult to
manipulate. In practice, electrode potentials on the hydrogen scale are usually
determined indirectly by measuring the e.m.f. of a cell formed from the electrode
in question and a convenient reference electrode whose potential with respect to
the hydrogen electrode is accurately known. The reference electrodes generally
used are the calomel electrode and the silver-silver chloride electrode (see
Sections XFV, 3-4).

When metals are arranged in the order of their standard electrode potentials,
the so-called electrochemical series of the metals is obtained. The greater the
negative value of the potential, the greater is the tendency of the metal to pass into
the ionic state. A metal will normally displace any other metal below it in the
series from solutions of its salts. Thus magnesium, aluminium, zinc, or iron will
displace copper from solutions of its salts; lead will displace copper, mercury, or
silver; copper will displace silver.

The standard electrode potential is a quantitative measure of the readiness of
the element to lose electrons. It is therefore a measure of the strength of the
element as a reducing agent in aqueous solution; the more negative the potential
of the element, the more powerful is its action as a reductant.

It must be emphasised that standard electrode potential values relate to an
equilibrium condition between the metal electrode and the solution. Potentials
determined under, or calculated for, such conditions are often referred to as
reversible electrode potentials, and it must be remembered that the Nemst
equation is only strictly applicable under such conditions.

n, 21. CONCENTRATION CELLS. An electrode potential varies with the
concentration of the ions in the solution. Hence two electrodes of the same metal,
but immersed in solutions containing different concentrations of its ions, may
form a cell. Such a cell is termed a concentration cell. The e.m.f. of the cell will be
•ence of the two potentials, if a salt bridge be inserted to
-liquid junction potential. It may be calculated as follows. At

the algebraic diffei
eliminate the liquid
25 °C:

„ /0.0591, „

^ Cl -t- - — - — log C 2 +

_ 0.0591 , Cl

log— , where c, > c,

n Cj i z

As an example we may consider the cell :

AgNO, aq.
[Ag+] = 0.00475M

AgNOjaq.
[Ag+] = 0.043M

Assuming that there is no potential difference at the liquid junction :

0.0591 . 0.043

■log

0.00475

= 0.056 volt

inte?p^» CALCULATION OF THE e.in.f. OF A VOLTAIC CELL. An
mg application of electrode potentials is to the cajculation.qf the e.m.f. of a

n, 23 QUANTITATIVE INORGANIC ANALYSIS

voltaic cell. One of the simplest of galvanic cells is the Danicll cell. It consists ofa
rod of zinc dipping into zinc sulphate solution and a strip of copper in copper
sulphate solution ; the two solutions arc generally separated by placing one inside
a porous pot and the other in the surrounding vessel. The cell may be represented
as:

ZnjZnS04aq.i!CuS0.,aq.iCu

At the zinc electrode, zinc ion.s pass into solution, leaving an equivalent negative
charge on the metal. Copper ions arc deposited at the copper electrode, rendering
it positively charged. By completing the externa! circuit, the current (electrons)
passes from the zJne to the copper. The chemical rc;iction.s in the cell are as
follows:

{a) zinc electrode, Zn ci Zn* ' + 2e;

(b) copper electrode, Ctr^ -t^cz^eCu

The net chemical reaction is:

Zn + Cu=- =Zn=*+Cu

The potential difference at each electrode may be ciilculatcd by the fomtula given
above, and the e.m.f. of the cell is the algebraic din’ercncc of the two potentials, the
correct sign being applied to each.

As an example we may calculate (he e.m.f. of the Danicll cell with molar
concentrations of zinc ions and copper (11) ions :

E = - Eg^ = + 0.34 - { - 0.76) =1.10 volts

Tlic small potential difference produced at the contact between the two solutions
(the so-called liquid-junction potential) is neglected.

II, 23. OXIDATION-REDUCTION CELl^. Reduction is accompanied by
a gain of electrons, and oxidation by a lo,ss of electrons. In a system containing
both an oxidising agent and its reduction product, there will be an equilibrium
between them and electrons. If an inert electrode, such as platinum, is placed in a
redoxsystem, for example, one containing rc(IIl)and Fc(n)ions,it willassumca
definite potential indicative of the position of equilibrium. If the system tends to
act as an oxidising agent, then Tc*' — Fc^ * and it will tahe electrons from the
platinum, leaving the latter positively charged; if, however, the sy.stcm has
reducing properties (Fc^^ — ►Fc’'*), electrons will be given up to the metal,
vvhich will then acquire a negative charge. Tlic magnitude of tlic potential will
t us be a measure of the oxidising or reducing properties of the system.

0 o tain comparative values of the ‘.strengths' of o.xidising agents, it is
neccssaiy, as in the ease of the electrode potentials of the metals, to measure under
standard experimental conditions the potential difference between the platinum
1"“ standard of reference. The primary standard is the

hydrogen electrode (Section 11, 20) and its potential is taken
whiVi, o, ‘‘rioard experimental conditions for tlic redox svstem arc those in

FUNDAMENTAL THEORETICAL PRINCIPLES H, 23

The potential measured in this way is called .the standard reduction potential. A
selection of standard reduction potentials is given in Table II, 4.

The standard potentials enable us to predict which ions will oxidise or reduce
other ions at unit activity (or molar concentration). The most powerful oxidising
agents are those at the upper end of the table, and the most powerful reducing

Table H, 4. Standard reduction potentials at 25°C

Halt-reaction £®, volts

Fj-f-2e;;i2F-
^2^8^ -f-2e^2S04^

Pb'^-l-2e:^Pb2 +

Mn04-+4H+-f3e?iMnO2-t-2H2O
Ce*'^+e;;iCe^''^ (nitrate medium)

BrOj- -h6H+ -i-5e:;iiBr2 +3H2O
Mn04 ' -I- 8H + -)- 5e Mn^ + -I- 4H2 O
Ce*'^ +e;^Ce’'^ (sulphate medium)
Ci2+2e:;±2Cl-

Cr20,^“ + 14H+4-6e:^2Cr^+ -f7H20
Tl^^-l-2e:^Tl+

Mn02-t-4H+ -t-2e;;iMn^+ -t-2HjO

Oj-)-4H++4e:;i2HjO

103--(-6H++5e?iiIj-(-3H2O

Brj-h2e^2Br-

HNOj+H^+ei-NO-t-HjO

NOj" -t-4H+ -|-3e^N0-)-2H,0

2Hg^"+2e;-Hg22^

C10'-|-H20-(-2e^±CI' -t-20H'

Cu^''^-(-I~-(-e:;iCuI

Hg2^+2e;;i2Hg

Fe^''+e;;iFe^+

BrO* +HjO-t-2e :;±Br~ -f 20H’
Br03'-|-3H20-F6e;;iBr~ -veOH'
Mn04^- + 2H2O -F 2e MnOj 40H "
Mn04--i-e:;iMn042-

H3ASO4 -F 2H + + 2e H, ASO3 -f H2O
Cu^" + Cl--Fe-CuCI
l2+2e;;i2r

JP' +H20-F2e:^r +20H"
+e-[Fe(CN)6f“
+‘’H+-F2e:;±U‘‘-^-F2H20
IO3 -F3H20-F6e^I-+6OH'

W^ + e;;iCu+

Sn‘*'''+2£^Sn^''‘

W++2H++e^TP++H20

SA -F2e:;±2S203^-

2H^-F2€-H,

Cp^-Fe;^Cr^ +

ZnSm PSn02]- -FH2O-F:

H aioi- t2«-Zn-F40H-
al +H20-F3e:;i:Al-F40H“

-F2.65
+ 2.01
+ 1.82
+ 1.70
+ 1.69
+ 1.61
+ 1.52
+ 1.52
+ 1.44
+ 1.36
+ 1.33
+ 1.25
+ 1.23
+ 1.23
+ 1.20
+ 1.07
+ 1.00
+0.96
+0.92
+0.89
+0.86
+0.79
+0.77
+0.76
+0.61
+0.60
+0.56
+0.56
+0.54
+ 0.54
+0.49
+0.36
+0.33
+0.26
+0.15
+0.15
+ 0.1
+0.08
0.00
-0,26
-0,41
-0.44
-0.56
-0.61
-0.67 _
-0.90 ;f
- 1-.22 '‘ •
-2.35

II, 24 QUANTITATIVE INORGANIC ANALYSIS

agents at the lower end. Thus permanganate ion can oxicIi.se Cl". Br". I", Fe^'‘
and (Fe(CN)6l'* " : ettn oxidise HjAsOj and I" but not Cr207^" or C!'. h

must be emphasised that for many oxidants the pFf of the medium is of great
importance, since tlicy are generally used in acidic media. Tims in mca.suring the
standard potential of the Mn04"-Mn’'’ system; MnO^' + SH* -f 5e = Mn’‘
+4H2O it is necessary to state that the hydrogen-ion activity is unity; this leads
to = +1 .52 volts. Similarly, the value of for the CrjO,' '‘-Cr’ * system is
-t- 1.33 volts. Tliis means that' the Mn04"“Mn^* system is a better o.xidising
agent than the CrjOT^'-Cr^'* sy.stcm. Since the standard potentials for Clj,
2Cr and Fe^‘-Fc^'‘ .systems' are + 1.36 and 0.77 volt respectively, per-
manganate and dichromate will oxidise Fefll) ions but only permanganate will
oxidise chloride ions; this explains why dichromalc but not permanganate
(except under very speci-.il conditions) can be used for the titration of Fefll) in
hydrochloric acid solution. Standard potentials do not give any information as to
the speed of the reaction: in some eases a catalyst is necessary in order that the
reaction may proceed with reasonable velocity.

Standard potentials are determined with full consideration of activity clfects,
and are really limiting values. They arc rarely, if ever, observed directly in a
potentiometric measurement. In practice, measured potentials determined under
defined conditions (formal potentials) are very useful for predicting the
possibilities of redox processes. Further details are given in Section X, 32.

11,24. CALCULATION OF THE STANDARD REDUCFION PO-
TENTIAL. A reversible oxidation- reduction system may be written in the
form [oxiilaiit = substance in oxidised state, rcductmi ~ .substance in reduced
state):

Oxidant -l-ne ;:i Rcductani
or Ox+m’:;±Rcd

The electrode potential which is established when an inert or unait.ickablc
electrode is immersed in a solution containing both oxidant and redtictant is
given by the expression:

where Ej- is the observed potential of the redox electrode at temperature T
relative to the standard or norma! hydrogen electrode taken as zero potential.
is the slmdard reduction potential.* 11 the number of electrons gained by the
oxidant in being converted into the rcductant, and a,^ and are theaclivitics of
the oxidant and rcductant rcsnociiv.'lv

they may be replaced by
of no great importance.

^P.nl

Since activities arc often difficult to determine directly
roncentrations; the error thereby introduced is usually
Itie equation therefore becomes:

e E of the oxidant and rcdiiciant. If boih activities arc variable,

e.g., he and Fe , corresponds to an activity ratio of unitv.

FUNDAMENTAL THEORETICAL PRINCIPLES D, 25

Substituting the known values of R and F, and changing from natural to common
logarithms, we have for a temperature of 25 “C (T = 298K);

0.0591

log

[Ox]

[Red]

If the concentrations (or, more accurately, the activities) of the oxidant and
reductant are equal, Eiy = E®, i.e., the standard reduction potential. It follows
from this expression that, for example, a ten-fold change in the ratio of the
concentrations of the oxidant to the reductant will produce a change in the
potential of the system of 0.0591/« volts.

n, 25. EQUILIBRIUM CONSTANTS OF OXIDATION-REDUCTION
REACTIONS. The general equation for the reaction at an oxidation-reduction
electrode may be written :

pA+qB+rC + iie:^sAr-t-ty-t-HZ+

The potential is given by ;

nF 4.uV-«z

where a refers to activiti'es, and n to the number of electrons involved in the
oxidation-reduction reaction. This expression reduces to the following for a
temperature of 25 °C (concentrations are substituted for activities to permit ease
of application in practice);

E = £^+^-.liog ■ Cc

p'i f,t

It is, of course, possible to calculate the influence of the change of concentration of
certain constituents of the system by the use of the latter equation. Consider, for
example, the permanganate reaction :

Mn04 -fSe^Mn^'*' -(-4H2O

(at 25 °C)

e concentration (or activity) of the water is taken as constant, since it is
assumed that the reaction takes place in dilute solution, and the concentration of
e water does not change appreciably as the result of the reaction. The equation

maybe written in the form;

£ = £®+2:2^1og[Mn04-

] 0.0591 , , ,8

-4— log[H+]®

[Mn^+] ' 5

calculate the effect of change in the ratio [Mn 04 “]/[Mn^'^] at
system concentration, other factors being maintained constant. In this

that th' ^’fficulties are experienced in the calculation owing to the fact

ion con products of the permanganate ion vary at different hydrogen-

mav Kn Rations. In other cases no such difficulties arise, and the calculation
cmp oyed with confidence. Thus in the reaction;

ir, 25 QUANTITATIVE INORGANIC ANALYSIS

H 3 ASO 4 4 - 2H ^ +2(’ HjAsOj + IfjO

£ = £® + -

[H 3 ASO 3

~t + j '“®|H,AsO,l Z

We arc now in a position to calculate the equilibrium constants of oxidation-
reduction reactions, and thus to determine whether such reactions can find
application in quantitative analysis. Let us consider first the simple reaction:

a,+2Fe=" 2?i2a'+2Fc^"

The equilibrium constant is given by ;

[CrfxlFc-^^]-

(CyxlFc-’^]-’

The reaction may be regarded as taking place in a voltaic cell, the two half-cells
being a Cl2,2Ci" .system and a Fc^'’,Fe^'‘ system. Tlie reaction is allowed to
proceed to equilibrium ; the total voltage or c.m.f. of the cell will then be zero, i-C.,
the potentials of the two electrodes will be equal ;

0.059 , (CL]

0.059, [Fc'"]

^ a,.:a d' 'og ~ ' »’ ’.r t‘ ' [pjTiTj

Now E'^a.-n ~ 1.36 volts and r,’- = 0.75 volt, hence

K = 4.7 X lO-"

The large value of the equilibrium constant signifies that the reaction will proceed
from left to right almost to completion, i.e., an ironlll) salt is almost completely
oxidised by chlorine.

Consider now the more complex reaction;

Mn04'-f5Fc^*+811’ ^Mn'’ +5Fc’" -l-4H,0
The equilibrium constant K is given by;

(Mn04 ■ ) X (Fc''* |fF f

The term 4 H 2 O is omitted, since the reaction is carried out in dilute solution, and
the water concentration may be assumed constant. The hydrogen-ion
concentration is taken as molar. The complete reaction may be divided into two
nall-ccll reactions corresponding to the partial equations;

Mn04--f8H^-f5c=iMn^’-f4H;0 0)

and

Fc^-" -i-f

For(l) as an oxidation-reduction electrode, wc have;

(2)

FUNDAMENTAL THEORETICAL PRINCIPLES H, 25

f 0-0^9, [MnO,-]x[H^r ■

t-t + ^ log .

0.059, [Mn04-]x[H+]®

^1.52-f— log—

The partial equation (2) may be multiplied by 5 in order to balance (1) electrically :

5Fe^+:;±5Fe3++5e ^2')

For (2') as an oxidation-reduction electrode:

Combining the two electrodes into a cell, the e.m.f. will be zero when equilibrium
IS attained, i.e.,

, _ 0.77+2^»,„g!£5!:i!

5 [Mn^^+J 5 ^[Fe^+js

= 63.5

or W [Mn^1x[Fe^+]^ 5(1.52-0^7) ^

[Mn 04 ]x[Fe^+]^ 0.059

[Mn 04 “] X [Fe^+J^ x [H+]® ^ ^ ^

that the reaction proceeds virtually to completion. It
case Thus sunn ^ ° the residual Fe(II) concentration in any particular

approximately^ O.liV-potassium permanganate with an
concentratio/nf h m iron(II) ' ions in the presence of molar

point be lOOcm^^hSp^+'i -Tm ofthe solution at the equivalence

practicallv nnn,J * r»i 7 + ■* T A, since it is known that the reaction is
exceS^^ [Mn ■^] = ix[Fe3-] = 0.002A, and [Fe^] = x. Let the

concentration point be one drop or 0.05 cm^; its

•‘-SS “f - »-V00 - 5 . .0-77 - tMnOr-!. Substtuting

^_(2xl0-a)x(ixl0--2)5

= 3 X 10«

or v =

^=[Fe'=+] = 5xlO-i5;V

stated that standard reduction potentials
for their possible ° pf^^^une whether redox reactions are sufficiently complete
ftat these calcubt'^^ quantitative analysis'. It must be emphasised, however,
opon which the nn°r* Pfovide no information as to the speed of the reaction,
jWs question mustV^^^^?^ iri practice will ultimately depend,

include the investi O', the basis of a separate experimental study, which may
the concentrations influence of temperature, variations of pH and of

theoretically notass'° reactants, and the influence of catalysts. Thus,
lum permanganate should quantitatively oxidise oxalic acid

11, 26 QUANTITATIVE INORGANIC ANALYSIS

in aqueous solution. It is found, however, that the reaction is extremely slow at
the ordinary temperature, but is more rapid at about 80 'C, and also incrctiscs tn
velocity when a little mangancse(ll) ion has been formed, the latter apparently-
acting as a catalyst. ,

It is of interest to consider the calculation of the equilibrium constant of the
general redox reaction, viz.:

a Ox, + h Red,, h Ox„ + o Red,

The complete reaction may be regarded as composed of two oxidation-reduction
electrodes. aOx„ a Red, and h 0.x, „ b Red,, combined together into a cell: at
equilibrium, the potentials of both electrodes are the same:

£, — £,

0.0591 , (Ox,]'

log

(Rcd,r

0.0591 ,

+ log

[ox„r

[Red,,]’'

At equilibrium. £, =£ 3 , hence;

This equation may be employed to compute the equilibrium constant of any
redox reaction, provided the two standard potentials £,'^ and £;^ arc known;
from the value of K thus obtained, the feasibility of the reaction in analysis may be
ascertained.

It can readily be shown that tlie concentrations at the equivalence point, when
equivalent quantities of the t wo substances Ox, and Red,, arc allowed to react, arc
given by:

[Ox,] [Red,,]

This expression enables us to calculate the exact concentration at the equivalence
point in any redox reaction of the general tvpe given above, and therefore the
feasibility of a titration in quantitative analysis.

II, 26. Selected bibliography

A full discussion of the topics considered in this Chapter wilt be found in
textbooks orphy.sical chemistry, and many of the topics are featured in textbooks
e ec roc icmistry-: a selection of books for further reading follows.

* Publi^iing'co^^^^ •5 o/hW// 7.)' oza//)// Calculations. Reading, Mass.; Addison Wesley

2. LK Butler (1964). Ionic Equilihrittm. Reading, M-ass.; Addison Wesley Publishing

3. C. W. Davies (1967). Electrochemistry. London; George Ncwncs Lid.

FUNDAMENTAL THEORETICAL PRINCIPLES U, 26

. 4 . R. B. Fischer and D. G. Peters (1969). A Brief Introduction to Quantitative Chemical
Analysis. 3rd edn. Philadelphia; W. B. Saunders Co.

5. H. A. Flaschka, A. J. Barnard and P. E. Sturrock (1969). Quantitative Analytical
Chemistry. Vol. I. New York; Barnes and Noble.

6. S. Glasstone (1942). Introduction to Electrochemistry. New York; Van Nostrand Co.
Inc.

7. L. F. Hamilton, S. G. Simpson and D. W. Ellis (1969). Calculations of Analytical
Chemistry. New York; McGraw-Hill Inc.

8. 1. M. Kolthoff and P. J. Elving (1959). Treatise on Analytical Chemistry. Part I,
Theory and Practice. Vol. I. New York; Interscience Publishers Inc.

9. H. A. Laitinen (1960). Chemical Analysis. An Advanced Text and Reference. New
York; McGraw-Hill & Co.

10. W. J. Moore (1972). Physical Chemistry. 5th edn. London ; Longmans.

IJ. F. Steel (1970). Grundlagen der Analytischen Chemie. 5th edn. Weinheim; Verlag
Chemie.

12. W. F. Sheehan (1970), Physical Chemistry. 2nd edn. Boston; Allyn and Bacon.

13. D. A. Skoog and D. M. West (1970). Fundamentals of Analytical Chemistry. 2nd edn.
London; Holt, Rinehart and Winston.

14. T. B. Smith (1940). Analytical Processes. 2nd edn. London; Arnold.

15. C. L. Wilson and D. W. Wilson (1959). Comprehensive Analytical Chemistry. Vol. lA.
Amsterdam; Elsevier Publishing Co.

16. J. G. Dick (1973). Analytical Chemistry. New York; McGraw-Hill Book Co.

CHAPTER III

COM! IWON APPARATUS
A£D BASIC TECHNIQUES

and Tte apparaius commonly

It IS essential that the bepinnc^r should r be described, and

fo ac,„i„ dcalcrily il Erd ,r 'tecproccdurcs.and

clean, orderly working must :ilsoV-cuI v-?, t^ f of apparatus. The habit of
points OTlI be helpful in this direction. <^bserv'ance of the following

any spillages ofsolid orhq^uKl'chcini^^^^^^ ‘^''‘'’''^ble so that

^ All glassware must be scr ml" ^ ^ immediately,

been standing for anv iencth of !ime n uisa T'" bas

water before use. The outs^dclf veskd! Iv'i ^'stilled or dc-ionised

which is resen ed cxclusivclv for this purnose -tn 1 r glass-cloth

cluttered with apparatuL All 'tll, become
openition should be grouped together 1 n

avoid confusion when duplicate fletcrniin .t' bench; this is most essential to

whichnofurthcrimmcdiatcusciscnvisic*'l Appanitus for

If lUvill be needed at a later st.i.. it nn'vl ^ returned to the locker, but
st. If a .solution, prccin. late filiri!i> '-'i *bc back of the bench.

He container must be .„adled so th it’the'l ''“bsequcni treatment,

the vessel must be suitably covered to '•'='« be readily identified, and

Hist : ,n this context, bark corks ire usm-ul of the contents bv

shed some du., For temporarybdel ne "'ey invariably tend to

i sl ’ r «" H glas.s is nrlr pencil or a felt tip pen

5 ReSem'S m"u",^ b.belli„g^l,;i;t5 '«

must be replaced on the riaglTsliKl? h” ‘’'f ""'"'"H on the bench; they
H It should be regarded nV n , '"’‘'""‘“''^''''■teru.se.

duplicate. ‘ that all determinations arc

provided for recording

flf tlerimnation, the title of whTcrlH °"'-lc page should be devoted to
ibsel, ''VO pages IL??"

Preccdurc bm ,J, fr“ “Id /or brief dcycripll of

5 g ""y special features associated

COMMON APPARATUS AND BASIC TECHNIQUES III, 2

mth the determination. In most cases it will be found convenient to divide the
page on which the experimental observations are to be recorded into two halves
by a vertical line ; the observation^ relating to duplicate determinations can then

be recorded side by side.

The record must conclude with the calculations and the results with
appropriate comments upon the degree of accuracy achieved.

BALANCES

in, 2. THE ANALYTICAL BALANCE. One of the most important tools of
the analytical chemist is the balance, and it is essential that the underlying
principles of the theory and construction of this fundamental item of equipment
should be understood.

The conventional free-swinging, equal-arm chemical balance is now rarely
used, but to appreciate the developments which have been incorporated in
present-day analytical balances (see Section HI, 6) it is necessary to discuss the

essential features of the simple balance. Such a
balance may be regarded as a rigid beam BC
having a central fulcrum (A) and two arras of
equal length; the two ends of the beam carry
prism edges upon which the balance pans are
supported by means of a suitable suspension
(Fig.III, 1). Let us suppose a body of mass Mi
is placed on the left-hand pan of the balance;
the pointer P (attached to the beam) will be
deflected to the right. To restore the pointer to
its original position, bodies of known mass,
termed ‘weights’, are added to the right-hand
d, the principle of the lever requires that the

k — d,

V A

-di—t

A/,

Fig. in, 1

pan. When equilibrium is restoi
following relation holds:

where f j and are the forces acting upon the left-hand and right-hand prism
^^P^ctively, and di and dj are the respective distances of these from the
the ori balance has equal arms, dj = d 2 and Fi = F^. Now,

the leff h” ^ forces F j and Ej lies in the attraction of gravity on the bodies in

and and right-hand pans respectively, or, otherwise expressed,

Fi = M^g and Ej = '

masses (or quantities of matter) in the left-hand and
speakinp ^^^P^^f^ely, and g is the acceleration due to gravity. Strictly
6) 1 and E 2 are the true weights in the two pans. But :

^2 ^ 2 g M2

gravity is en forces with which the bodies in the two pans are attracted by
interested f'''° masses. In quantitative analysis we are

y m the amount of matter in the body, i.e., in its mass: this is

COMMON APPARATUS AND BASIC TECHNIQUES ffl, 5

arms of the beam are of equal length, (ii) the beam is rigid and does not bend
appreciably under load, (iii) the three bearing edges lie in the same plane and are
parallel to each other.

(b) The balance must be stable, that is, the beam must return to the horizontal
position after swinging. This is attained by proper adjustment of the centre of
gravity.

[c) The balance must be sensitive, that is, 0.1 mg should be readily detectable
with average loads. We may define the sensitivity of a balance as the angular
defieetion a of the beam when a known small weight is added. It can be shown
that the angle a is determined by the excess of weight iv producing the deflection
a, the length, d, of the balance arm, the weight, W, of the beam, and the distance,
h, between the centre of gravity and the point of support of the beam. Expressed
mathematically (see Ref. 1), .

. wd

The angular deflection of the beam is equal to the angular deflection of the
pointer, and the latter is directly proportional to the number of divisions between
the two points of rest on the scale at the foot of the beam. This leads directly to the
usual definition of sensitivity, viz., the sensitivity of a balance is the number of
scale divisions that the rest point (of equilibrium point) is displaced by a weight
of 1 mg.

In an ideal balance, free from friction and with a perfectly rigid beam, the
sensitivity would be independent of the load. Most balances, however, exhibit a
decreasing sensitivity with increasing load, and this change of sensitivity provides
a pod criterion as to the maximum safe load that a balance can carry. The
cnterion is ; no greater load should ever be placed upon the balance pans than the
load at which the sensitivity becomes 40 per cent of its maximum value.

m, 5. WEIGHTS, REFERENCE MASSES. The determination of the mass of
an 0 ject with an equal-arm balance necessitates the use of a series of reference
masses termed weights. For scientific work, the international metric system of
eig ts and measures is employed. The fundamental standard of mass is the
in prototype kilogram, which is a mass of platinum-iridium alloy made

Pari ^^*6 International Bureau of Weights and Measures near

auth copies of the standard are kept by the appropriate responsible

the various countries of the world ; these copies are employed for

weieht^f secondary standards, which are used in the calibration of

in lahn "'o^k. The unit of mass that is almost universally employed

thou^anlk*^ work, however, is the gram, which rriay be defined as the one-
An nrH' international prototype kilogram.

30 20 of analytical weights contains the following: grams, 100, 50,

sequence Otv. ’ “iPigrams, 500-100 and 50-10 in the same 5, 3, 2, 1
but the nrn • sequences, such as 5, 2, 1, I, or 5, 2, 2, 1, are also encountered,
'vill also of duplicate weights is not recommended. The set of weights

am a rider (see Section in, 3) for weights below 10 mg. The weights

' (NPL) in Great Britain, the National Bureau of Standards

. COMMON APPARATUS AND BASIC TECHNIQUES ffl, 6

enlarged, to read up to 1 00 mg or even’up to 1 000 mg with consequent elimination
of some or all the fractional weights; in such cases some form of vernier or
micrometer is included to facilitate the reading of the fourth decimal place
(0.1 mg). The sensitivity of the balance is unaffected by the damping effect but
clearly the optical scale will only be accurate for loads within the range in which
the'sensitivity is constant.

B. Dial-controlled weight loading. Instead of having to transfer weights
individually from the box to the balance pan, it is a great improvement to have
the weights suspended within the balance case in such a position that a series of
hooked levers can place theih on (or remove them from) a carrier bar attached to
the right-hand arm of the balance beam. For ease in handling, the weights are
fabricated in ring or other convenient form, and the operating levers are
controlled by knobs with engraved dials on the outside of the balance case : these
dials are rotated to add or remove appropriate rings from the carrier bar. A
common arrangement is to provide two dials, one covering the range 0. 1-0.9 g
and the other the range 1-9 g : the weights from 1 0 g upwards are contained in a
special holder located within the balance case, and of course, the weight below
0.1 g is determined from an optical scale as described under A. Alternatively, the
optical scale may be used to cover the range 0-1000 mg and the two weight-
loading dials are then used respectively for the 1-9 g and 10-90g ranges:
additional weights are then no longer required.

C. Controlled release mechanisms. In order to maintain the sensitivity-load
relationship and the precision of weighing, wear on the knife edges of the balance
must be kept to a minimum. This is achieved by controlling the rate of impact of
the knife edges upon the supporting planes when the balance is released. In the
Releas-o-matic’ device used by Messrs Oertling, the beam movement was
controlled pneumatically by a graphite piston moving within a graphite cylinder :

,e use of graphite obviates lubrication problems and the device always operates
smoothly. The rate of movement of the piston is controlled by the rate at which
air enters the cylinder through an adjustable needle valve. Some manufacturers
use a small synchronous motor coupled to a gear unit which releases the beam at
a steady, slow rate.

jj- , ■ ^’’P^^lghing devices. The weighing operation can be greatly speeded up
Acc^(f^*i object to be weighed is already known approximately.

weieV ' ’ manufacturers now incorporate in their balances a ‘pre-
direct'*^^ farility. In some systems, the beam release lever is turned in one
until ° beam is only partially released, and weights are then added

positin ^ almost balanced. The release lever is then turned back to the rest
theoDtin ®till moving in the same direction, to the completely free position;

WithO ^ comes into operation and the full weight is obtained,
the beam Sartorius balances, operation of the pre-weigh lever leads to

system supported on a spring which is fitted with an optical read-out

weights approximate weight of the object (see Fig. Ill, 4). The correct

obtainpa selected on the weight-loading dials and the final weight is

In r f-r® i*"'

‘^orporated balance the foregoing improvements are in-

halance nan” ^t^anstrument in which the balance beam is unsymmetrical, one
operated w suspension is replaced by a counterpoise, and the dial-

supDort-tK^* , suspended from a carrier attached to the remaining pan
PPort. these changes are illustrated in Fig. Ill, 2.

Ill, 6 QUANTITATIVE INORGANIC ANALYSIS

from 1 g upwards arc constructed from a non-magnctic nickel - chromium alloy
(80% Ni, 20% Cr). or from austenitic stainless steel: plated brass is sometimes
used "but' is less satisfactory. The fractional weights arc made from the same
alloys, or from a non-tiirnishablc metal such as gold or platinum. For handling
the weights, a pair of forcep.s, preferably ivory-tipped, arc provided and the
weights arc stored in a box with suitably shaped compartments.

A set of weights should be calibrated before it is adopted for laboratoi^' use
and it is usually advisable to recalibrate at yearly intenmls: the calibration is
carried out by comparison of c.ich piece of the set against I he corresponding piece
from a standard set which has been calibrated by a nationally recognised
institution, c.g.. in Great Britain, the National Phy.sical l.aboratory. If a set of
standard weights is not available, then the weights in one set may be inter-
calibrated ; for details of this procedure .see T. W. Richards (Ref. .2).

The National Physical Lalioratory at Teddington recognises only onegrade of
weights, 'Class A’, in which I he following tolerances are permitted : 1 00 g. 0,5 mg;
50g, 0.25mg: 30g. O.lSmg; 2flg. O.lOmg; lOg-lOOmg, 0.05mg: 50-t0mg.
0.02 mg.

The National Bureau of Standards at Washington recognises the following
classes of precision weights;

C/o.t.s M. For use :ts reference standards, for work of the highest precision,
and where a high degree of constancy over a period of time is required.

Class S. For use as working reference .standards or as high precision
analytical weights.

Class S-!. Precision analyiicai weights for routine analytical work.

Class J. Microweight standards for microbalnncc.s.

It must be emphasised again (cf. Section 111, 2) th.nt the ‘weights’ which have
been discussed arc strictly masses; some laboratory suppliers do now list ‘boxes
of masses’, but most analysts will undoubtedly still refer to ‘weights’.

111,6. TWO-KNIFE SINGLF^PAN BALANCE. The simple balance
described in Section 111,2 h;ts Irccn subject to many modifications; these have
been chiclly directed towards improving tiic sjxicd of weighing, but other
beneficial results have also accrued. .An intcrcsiinc account of the development of
the analytical balance has been given by J. T. Stock (Ref. 2). Important
developments culminating in the introduction of tlic two-knifb single-pan
balance by E. Mcttlcr in 1946 arc discussed briefly in the following paragraphs.

A. Aperiodic balances. In the apcnodic balance an air-damping device is
attached to the beam : this consists of pi.stons attached to the beam and operating
m stationary cylinders closed at one end and with llic minimum clearance
etween piston and cylinder. The resulting damping eficct brings the beam to rest
(K ~ ^ seconds alter release, and if the object to be weighed is slightly hoarier than
c weights deployed in the pan, the beam will be tilted up, and the pointer will be
e ec e to the right. The displacement from the null position can be measured
(hl*r graticule attached either to the right-hand end of the beam, or to
or. 1 ° illuminated by a suitably placed lamp. The scale in the

of calibrated to measure the weight corresponding to the dcficction

o' arrangement the ridcrcan be dispensed with, and all

^ 10 mg can be read on the optical scale: an appropriate optica!

orr, racorporated so that the scale reading is displavcd on a suitably placed
ground-glass screen. The principle can be extended.' and the optical scale

Ill, 8 QUANTITATIVE INORGANIC ANALYSIS

up to about I kg to be weighed rapidly with an accuracy oF about O.I g, the top
loading balance has now become an indispensable item oflaboratory equipment.
Slightly dilFercnt models \GI1 cater for loads of up to 5 kg with an accuracy of
about b.I g on the one hand, or with loads of up to 200 g with an accuracy of
0.0 1 g or even 0.001 g: with thc.se more sensitive models it is c.sscnttal that the bal-
ance be presided with some form of screen to shield the pan from draughts. A
typical balance of this type (Ocrtling Model TD.^0) is shown in Fig. 111. 6.

FiR.ni,6

In such a balance weighing up to 1 kg. weights moving up in steps of lOOg are
controlled by a knob at the side of the balance, and the weight selected appears as
a digit on an optical read-out panel on the front of the instrument. The remaining
weights (tens, units and tenths of a gram) then appear on the same read-out panel
and are obtained from a magnified image of a graticule attached to the balance
beam. Such balances incorporate a special knife suspension system, and the
speed of weighing is partly attained with the aid of a magnetic damping device to
reduce the oscillations of the beam. A taring device is frequently fitted, and this is
very useful when it is required to weigh out a given quantity of material.

in, 8. ELECTRONIC BALANCES. A recent development in the design of
two-knife single pan balances is the replacement of the optical read-out system
used for the fractional weights by an electrically operated measuring system. The
principle of the system incorporated in the Mcttlcr HE20 electronic analytical

COMMON APPARATUS AND BASIC TECHNIQUES III, 9

balance is shown in Fig. Ill, 7. Dial-operated weights (tens and units) are added
or removed in accordance with the usual substitution principle; the weights
selected are detected electrically (9) and the appropriate figures displayed on a

digital display unit (10). The fractional part of the weight is determined with the
aid of an electrical sensor (3) which responds to the deflection of the beam, and
modulates an amplifier (5), thus generating a current in the ‘compensation
system (2). The compensation system current gives rise to a magnetic field which
serves to maintain the beam in equilibrium: the magnitude of the current
necessary to produce the requisite magnetic field is proportional to the weight
wtuchis being determined. This information is likewise transmitted to the display
ni ). Connection (12) can also be made to a print-out unit so that a
^ f weight can be recorded automatically,

novel form of top-loading balance introduced by
object nlarorf A Ltd It contains no moving parts, and the weight of an

conjunctionwifha^^rH^l^^ balance is obtained by means of a transducer in
tube digital display electncal circuit ; the weight is indicated on a neon-

Amongst other types of balances,

is replaced of ^bis type, the normal beam

“etal bands which are nulledT parallel beams, held in position by

employed : one at each end supports. Three supports are

and one in the centre wVi.Vu assembly which carry the balance pans,

edges to wear, and as the complete assembly. There are no knife

up. ’ <^065 not have to be arrested, weighing is speeded

horizontal beTm^atta"ched toh^. fine wire or fibre is stretched taut, and a
he beam, thus pulling it downwn if a weight is placed at one end of

"'hich causes it to twist Bv mp suspension wire experiences a torque

J’^Pension wire, the lattw cTn h? graduated dial attached to one end of the
he horizontal position. The dint manually until the beam is returned to

Tfe directly calibrated against known weights so that

0-0001 mg. IS made from a quartz fibre, and which weighs to

L- ^tectrobalanrPf tu - . .

s erm is applied to balances such as the Cahn

Ill, 8 QUANTITATIVE INORGANIC ANALYSIS

up to about 1 kg to be weighed rapidly with an accuracy of about 0 1 c the

loading balance has now become an indispensable item of l-ihor-.tnr,..’,, ^

Shghtly dilTerent models will cater for loads c up to

about 0.1 g on the one hand, or with loads of up to ->00 n w

0.01 goreven 0.001 g: with , he.se moresensiti^ models- ^

ance be provKlcd with some form of screen to shield the n^n from T

ti'pical balance of this type (Oertling Model TD30) i.s shoum in^

^ 18 . 111,6

In such a balance wcichinn nr, » i i

controlled by a knob at the side SfUicbalm?''*’' w'''"®

'^■>0-out panel on tim r ' '^"r ' appears as

“"its and tenths of a enn fr f™"' ‘ The rcrnaminc

from a magnified im !' ,? "PP"*'!''®" 'Sesame read-out panel
r“‘' ' '"''“rporaie JoS- Pi «l'achcd to the balance

rL^^°^'''‘='Sking is partly attained whP.n P” suspension .system, and the
reduce the oscillations of the beam A t- ' ‘‘ '"•’’emetic damping device to

m. 8 , ELECTRONIC

used'fr?i.‘'''f^'^ P'^" ‘l^vclopmcnt in the design of

P nciplc of the system incorpoJated n t h m measuring .system. The

66 Hn20 electronic analytical

COMMON APPARATUS AND BASIC TECHNIQUES HI, 9

balance is shown in Fig; III, 7. Dial-operated weights (tens and units) are added
or removed in accordance with the usual substitution principle; the weights
selected are detected electrically (9) and the appropriate figures displayed on a

digital display unit (10). The fractional part of the weight is determined with the
aid of an electrical sensor (3) which responds to the deflection of the beam, and
modulates an amplifier (5), thus generating a current in the ‘compensation
system (2). The compensation system current gives rise to a magnetic field which
serves to maintain the beam in equilibrium: the magnitude of the current
necessary to produce the requisite magnetic field is proportional to the weight
nnjf determined. This information is likewise transmitted to the display

nprmin ^ print-out unit so that a

P rmanent record of the weight can be recorded automatically.

Internafinnalm^^t is a novel form of top-loading balance introduced by

object dared

conjunction with a cni'H is obtained by means of a transducer in

tube digital display ' electneal circuit ; the weight is indicated on a neon-

Amegs, other types of balances,

is replaced b^two^comrafrhTr'^'i of this type, the normal beam

metal bands which are m.lli t V ^ parallel beams, held in position by
employed: one at each end nf th supports. Three supports are
and one in the centre which suonort^Ih '^hich carry the balance pans,

«dges to wear, and as the him H assembly. There are no knife

up ^ the beam does not have to be arrested, weighing is speeded

JorizontalSa^a^^^^^^ fine wire or fibre is stretched taut, and a

he bean,_ thus pulling it downwairjh ’

‘^^'^sosittotwisl By means n A wire experiences a torque

“ pension wire, the latter cSi h Ant of the

it position. The dial can ™®"^nlly until the beam is returned to

Thk iu fifiectly calibrated against known weights so that

0.0001nig IS made from a quartz fibre, and which weighs to

'^^^^obalances Tliio * •

term is applied to balances such as the Calm

Ill, 10 QUANTlTATIvn INORGANIC ANALYSIS

electrobaliincc, in whicli an electromagnetic force is used to counteract the
dellcction of the balance beam caused by adding a weight to one side. At the
fulcrum of the beam, and at right angles to it. a wire coil is attached. This coil is
mounted between tlic poles of a permanent magnet, one situated above, and the
other below llic coil. If a current is passed through tlic coil the resultant
electromagnetic interaction applies a torque to the beam. The current is adjusted
by means of a potentiometer until the beam is rc.stored to the null position, the
potentiometer dial being calibrated to read weights directly: a sensitivity of

0. 1-0.02 /ig can be achieved.

Ill, 10. CARE AND USE OF ANALYTICAL BALANCMS. No matter what
type of analytical balance is employed (single pan or two pan), due attention
must be paid to the manner in whicli it is used. The following points should be
carefully observed.

1. The balance should be placed upon n firm foundation wliich is as free from
mcchanic,nl vibration as possible. The ideal foundation is a concrete orstoncslnb
resting upon brick piers, which arc either sunk into the ground or. if tins is not
practicable, into the concrete floor or sub-floor of the laboratory, if this i.s not
possible, the balance should be set up on a stout table or shelf and protected,
when necessary, by sheets of shock-absorhing media, such ns cork mats or sheet
rubber on which the balance is placed: anti-\ibration tablc-s, dc,signcd for
balances, are available from most laboratorj suppliers. It is best to keep the
balance in a room separate from the laboratory in order to protect it from fumc.s.
and it sliould be located in a draught-free position away from direct sunlight.

The balance must be level. This luljuslmcnt may be made with the aid of the
levciling-scrcws and spirit levels on the base of the instrument.

2. When not in use. the balance beam should be raised so as to protect the
knife edges and bearing planes. The doors of the balance should be kept clo.sed
whenever possible.

i. To release the balance, the beam should be lowered very gently.

4. Objects to be weighed must be allowetl to attain the temperature of the
balance before weighing is attempted, otherwise tlic air currents produced inside
the balance case may introduce errors. If the object has been heated, svifhcicnt
time must be allowed for cooling. The lime required to .attain the balance-room
temperature varies with lire si'/c, etc., of the object, but as a rule 30-40 minutes is
suflicient.

5. The object to be wciglicd should always be placed in the centre of the pan ;
the same remark applic.s to the weights if a two pan balance is used.

6. 1 he weights used with a non-dial loadina balance must be handled only
with the forceps provided.

7. When objects arc being added to or removed from the pan, tite beam arrest
must be raised so as to protect the knife cdgc.s from injury. A similar remark
applies to alteration of the weights on a single pan balance, unless it is in the ‘pre-
weigh’ position.

S. As soon as all external weights have been added, the balance ease must be
closed. Hence with a fully dial-operated single pan balance, tlie ease will be closed
as soon as the object has been placed on the pan.

9. No chemicals or objects which might injure the balance pans should ever be
placed directly upon them. Substances must be weighed in suitable containers,
such as small beakers, weighing bottles or crucibles, or upon watch glasses.

COMMON APPARATUS AND BASIC TECHNIQUES III, 11

Liquids and volatile or hygroscopic solids must be weighed in tightly closed

vessels, such as stoppered weighing bottles ^ ^

10. The balance must not be overloaded (see Section III, 4 (c) j.
n. Nothing must be left on the panwhen the weighing has been completed. If
anysubstance is spilled accidentally upon the pan or upon the floor of the balance
case, it must be removed at oncei The pans should be lightly brushed periodically
with a camel-hair brush to remove dust which may have collected.

12. A beginner should never attempt to adjust a balance: help should be
sought from an experienced operator.

In the actual weighing process, the exact sequence of operations w'ill be partly
dependent upon the make of balance in use and the arrangement of the controls,
but with a single pan balance it will include the following steps.

1. Sit opposite the centre of the balance.

2. Brush the pan lightly with the camel hairbrush to remove any dust.

3. Carefully release the beam and check that the empty balance gives a zero
reading: if necessary the requisite adjustment should be made.

4., With the beam at rest, place the object, which must be at or near room
temperature, on the pan, and close the balance case.

5. Set the balance to the ‘preweigh’ position and from the scale reading select the
appropriate gram weights with the weight loading dials.

6. Release the beam fully and record the final weight: with some balances this
may necessitate the adjustment of a vernier control to enable the fourth place
(0.1 mg) to be read.

7. When weighings are completed, arrest the beam, return the weight dials to
zero, remove the object which has been weighed, clear up any accidental
spillages, and close the balance case.

1. chief sources of error arc the following:

successive weighingr°'^ of the containing vessel or of the substance between

3’ ScemijSe

1. The first •

-^vuidey or tne weights.

vessel; (a) bv occasioned by change in weight of the containing

caused by rubhinJ^«lTt°s u • ’Moisture, (b) by electrification of the surface
balance case. These e ^ temperature being difierent from that of the
Mth a linen cloth may be largely eliminated by wiping the vessel gently

before weighing The ele stand at least 30 minutes in the balance room

P‘>i‘ticularly if both thp "t ” ’‘'7»'c/r may cause a comparatively large error,

standing; it may be are dry, is slowly dissipated on

radioactive material in 'th k ^ P*oce of pitchblende or similar feebly

efflorescent, and volatile ^ the air. Hygroscopic,

vessels. Substances which h ^ost be weighed in completely closed

®ra generally allnwpH heated in an air oven or ignited in a crucible

’j. _ 1. f _ J ^

- temperature and oannot be exactly specified, since it will depc —

raaterial of which it is mm the crucible as well as upon the

ose of porcelain slass « ' ^^^tinum vessels require a shorter time than

raicibles in the desiccator for Pn 'os* customary to leave platinum

-25 minutes, and crucibles of other materials for

in, 11 OUANTITATIVR INORGANIC ANALYSIS

30-35 minuies before being weighed. It is advisable to cover crucibles and other
open vessels.

2. When a substance is immersed in a fluid, il.s true weight is diminished by the
•weight of the fluid which it displaces. If the object and the sveights have the same
density, and consequently the same volume, no error will be introduced on this
account. If, however, as is usually the case, the density of the object is different
from that of the weights, the volumes of air displaced by each will be different. If
the substance has a lower density than the sveights, as is usual in analysis, the
former will displace a greater volume of air than the latter, and it will therefore
weigh less in air than in a vacuum. Conversely, if a denser material (c.g., one of
the precious metals) is weighed, the wciglu in a vacuum will be less than the
apparent weight in air.

Consider the weighing of I litre of w.atcr. first in vacuo, and then in air. It is
assumed that the flask containing the water is tared by an exactly similar flask,
that the temperature of the air is 20 'C and the barometric pressure is 760 mm of
mercury. Tiie weight of 1 litre of water in vacuo at 20 C and 760 mm is 998.23 g. If
the water is weighed in air, it will be found that 998.23 g arc too heavy. We can
readily calculate the difference. The weight of 1 litre of air di.splaced by the water
is l.lOg. Assuming the weights to have a dcnsiiv of 8.0. thev will displace
998.23/8.0 = 124.8. or 124.8 x 1.20/1000 =■- 0.15g of air. The net diflcrcncc in
weight will therefore be 1.20—0.15 = 1.05 g. lienee the weight in air of 1 litre of
water under the experimental conditions named is 998,23- 1.05 = 997. j 8 g, a
difference of 0. 1 per cent from the weight in vacuo.

Ia:t us now extend our enquiry to the ease of a solid, such as potassium
chloride, under the above conditions. The density of potassium chloride is 1.99.
If 2g of the salt are weighed, the apparent lo.ss in W’ciglil (= weight of air
displaced) is 2x0.0012/1.99 = 0.0012g. The apparent loss in weight for the
w'cights i.s 2 X 0.0012/8.0 = 0.00030 g. Hence 2 gofpotassium chloride will weigh
0,0012 — 0.00030 = 0.00090 g less in air than in vacuo, a difference of 0.05 per
cent.

COMMON. APPARATUS AND BASIC TECHNIQUES ffl, 12

Since the difference between and does not usually exceed, 1 to 2 parts per

thousand, we may write: ’

where . : ■

1.20(^-g-Q)

The values of k for = 0.0012 and ~ 8.0 have been calculated and are
collected in Table III, 1. If a substance of density d^ weighs grams in air, then
W^.k milligrams are to be added to the weight in air in order to obtain the weight
in vacuo. The correction is positive if the substance has a density lower than 8.0
(stainless steel), and negative if the density of the substance is greater than 8.0.

Table HI, 1 Reductions of weighings made in air

mth weights of density 8.0 to vacuo

0 !

+ 2.25

1.9

+0.48

11.0

lilM

+ 1.85

2.0

Bixfn

0.7

+ 1.56

2.5

+ 0.33

13.0

-0.06

0.8

+ 1.35

3.0

+0.25

14.0

-0.06

0.9

+ 1.18

3.5

+0.19

15.0

-0.07

1.0

+ 1.05

4.0

+ 0.15

16.0

-0.07

1.1

+0.94

4.5

+0.12

17.0

-0.08

1.1

+0.85

5.0

+0.09

18.0

-0.08

1.3

+0.77

5.5

+0.07

19.0

-0.09.

1.4

+0.71

6.0

+0.05

20.0

-0.09

1.5

+0.65

7.0

+0.02

21.0

-0.10

1.6

+0.60

8.0

+0.00

22.0

1.7

+0.56

9.0

-0.02

23.0

1.8

+ 0.52

10.0

-0.03

24.0

3. Accuracy of the weights can be ensured by periodical checks against, a
standard set of weights.

GRADUATED.GLASSWARE

in, 12, UNITS OF VOLUME. For scientific purposes the convenient unit to
employ for measuring reasonably large volumes of liquids is the cubic decimetre
(dm ), or, for smaller volumes, the cubic centimetre (cm^). For many years the
fundamental unit employed was the litre, based upon the volume occupied by
one kilogram of water at 4 °C (the temperature of maximum density of water) ;
the relationship between the litre as thus defined and the cubic decimetre was
established as ■ ■

1 litre = 1.000028 dm^
or 1 millilitre = 1 .000 028 cm^

COMMON APPARATUS AND BASIC TECHNIQUES m, 14

Table III, 3 Temperature corrections for volumes of water
measured in a 1000-cm^ glass flask (standard temperature, 20°C)

Temperature Correction (cm^)

Soda glass

Borosilicate glass

+1.37

+ 1.61

+ 1.24

+ 1.40

+0.77

+0.84

20 ,

0.00

-1.03

-1.11

.-2.31

-2.46

III, 14. GRADUATED APPARATUS. The most commonly used apparatus in
titrimetric (volumetric) analysis are graduated flasks, burettes, and pipettes.
Graduated cylinders and weight pipettes are less widely employed. Each of these
will be described in turn .

Graduated apparatus for quantitative analysis is generally made to specifi-
cation limits, particularly with regard to the accuracy of calibration. In Great
Britain there are two grades of apparatus available, designated Class A and Class
B by the British Standards Institution. The tolerance limits are closer for Class A
apparatus, and such apparatus is intended for use in work of the highest
accuracy: Class B apparatus is employed in routine work. In the United States,
specifications for only one grade are available from the National Bureau of
Standards at Washington, and these are equivalent to the British Class A.

Qeaning of glass apparatus. Before describing graduated apparatus in
detail, reference must be made to the important fact that all such glassware
must be perfectly clean and free from grease, otherwise the results will be
unreliable. One test for cleanliness of glass apparatus is that on being filled with
distilled water and the water withdrawn, only an unbroken film of water remains.
If the water collects in drops, the vessel is dirty and must be cleaned. Various
methods are available for cleaning glassware.

Many commercially available detergents are suitable for this purpose, and
some manufacturers market special formulations for cleaning laboratory
glassware; some of these, e.g., ‘Decon 90’ made by Decon Laboratories of
ortslade, are claimed to be specially effective in removing contamination due to

radioactive materials.

Teepol’ is a relatively mild and inexpensive detergent which may be used for
ceatung glassware. The laboratory stock solution may consist of a 10 per cent
so ution in distilled water. For cleaning a burette, 2cm^ of the stock solution
j. with 40 cm^ of distilled water is poured into the burette, allowed to stand
a° i? ^ the detergent run off, the burette rinsed thrice with tap water,

n then several times with distilled water. A 25 cm^ pipette may be similarly

eaned using 1 cm^ of the stock solution diluted with 25-30 cm^ of distilled
Water.

‘cl^ ?®thod which is frequently used consists in filling the apparatus with
dich ™^Iure’, a nearly saturated solution of powdered sodium or potassium

hou concentrated sulphuric acid, and allowing it to stand for several

thor^’ ?^^^®.^^hly overnight; the acid is then poured off, the apparatus
oughly rinsed with distilled water, and allowed to drain until dry. [It may be

Ill, 15 QUANTITATIVE INORGANIC ANALYSIS

mentioned that potassium dichromatc is not very soluble in concentrated
sulphuric acid (about 5p per litre), whereas sodium dichromate
Na 2 Crj 07 . 2 H ,0 is much more soluble (about 70g per litre); for this reason, as
well as the fact that it is much cheaper, the latter is usually preferred for the
preparation of ‘cleaning mixture’. From time to time it is advisable to filter the
sodium dichromate -sulphuric acid mixture through a little glass wool placed in
the apex ofa glass funnel: small particles or sludge, which arc often present and
may block the tips of burettes, arc thus removed.] A more cfiicicnt cleaning liquid
is a mixture of concentrated sulphuric acid and fuming nitric acid; this may be
used if the vessel is very greasy and dirty, but must be handled with extreme
caution.

A very effective de-greasing agent, which it is claimed is much quicker tiding
than ‘cleaning mixture’, is obtained by dissolving lOOg of potassium hydroxide
in 50cm^ of water, and after cooling, making up to I litre with industrial
methylated spirit (Ref. fib).

HI, 15. GRADU.A’FFD FLASKS. A graduated llask (known alternatively as a
volumetric flask or a measuring flask), is a llat-bottomcd, pear-shaped vessel with
;i long narrow neck. A thin line etched around the neck indicates the volume that
it holds at a certain definite temperature, usually 20 X (both the capacity and
temperature arc clearly marked on the flask): the flask is then said to be
graduated to contoin. Flasks with one mark are always taken to amiaiit the
volume specified. .A flask may also be marked it> iMhrr a specified volume of
liquid under certain definite conditions; tiicsc arc, however, not suitable forc.xnct
work and arc not widely used. Vessels intended to contain definite volumes of
liquid arc marked C or TC or In. while those intended to deliver definite volumes
arc marked D or TD.

The mark extends completely around the neck in order to avoid errors due to
parallax when making the final adjustment: the lower edge of the meniscus
should be tangential to the graduation mark, and botli the front and the back of
the mark should be seen as a single line. The neck is made narrow so that a small
change in volume will liave a large clfcct upon the height of the meniscus: the
error in adjustment of the meniscus is accordinglv small.

The flasks should be fabricated in accordance with BS 1792 and the opening
should be ground to standard (interchangeable) specifications and fitted wath an
interchangeable glass or plastic (commonly polypropylene) stopper. They should
conform to either Class A or Class B specification; examples of permitted
tolerances for the latter Grade arc as follows ;

Flask size 5 25 100 250 1000 env’'

Tolerance 0.04 0.06 0.15 0.30 O.SOcnr’

For Class A flasks the tolerances arc appro.ximatcly halved; such flasks may be
calibration certificate, or witli a British Standard Test

(BST) Certificate.

available in the following capacities: 1, 2, 5, 10. 20, 50.
2000 and 5000cm^. They arc employed in making Up
sandard solutions to a given volume; they can also be used for obtaining, with
^ of pipettes, aliquot portions of a solution of the substance to be analysed.

Caubrabon. For most analytical purposes flasks of Class A standard may
e use without calibration, but for the highest accuracy, all flasks (unless

COMMON APPARATUS AND BASIC TECHNIQUES IH, 15

carrying a recent BST Certificate) should be calibrated ; this involves determining
the weight of water held by the flask when it is filled to the mark. For this purpose
a large balance which will accommodate the largest flask to be calibrated (say one
litre) is required: a top pan balance of suitable loading and sensitivity
characteristics maybe used. ■ . •

The flask is first thoroughly cleaned and dried, and after standing in the
balance room for an hour is stoppered and weighed. A small filter funnel, the
stem of which has been drawn out so that it reaches below the graduation mark of
the flask, is then inserted into the neck and de-ionised (distilled) water, which has
also been standing in the balance room for an hour, is added slowly until the

Table III, 4 Weight of water to give one litre at 20 °C*

Flask of soda glass, coefficient of cubical expansion, O.OOOU25/°C

Temp.

Weight

Volume of

Temp.

Weight

Volume of

rc)

(g)

1 g of water

(“O

(g)

1 g of water

(cm^)

(cm’)

998.39

1.0016

996.60

1.0034

998.32

1.0017

996.38

1.0036

12 '

998.23

1.0018

996.17

1.00385

998.14

I.OOI85

995.93

1.0041

998.04

1.0019

995.69

1.0043

997.93

1.0021

995.44

1.0046

997.80

1.0022

995.18

1.0048

997.66

1.0023

994.91

1.0051

997.51

1.0025

31 .

994.64

1.0054

997.35

1.0026

994.35

1.0057

997.18

1.0028

994.06

1.0060

997.00

1.0030

993.75

1.0063

996.80

1.0032

993.45

1.0066

Flask of borosilicate glass, coefficient of cubical expansion, 0.000010/°C

Temp.

(T)

Weight

Volume of

Temp.

Weight

Volume of

(g)

1 g of water

(°C)

(g)

1 g of water

(cm’)

998.00

1.0020

996.33

1.0037

997.86

1.0021

996.09

1.0039

997.71

1.0023

995.85

1,0042

997.54

1.0025

995.49

1.0045

997.37

1.0026

995.32

1.0047

997.18

1.0028

995.05

1.0050

996.98

1.0030

994.76

1.0053

996.78

1.0032

994.47

1.0056

996.56

1.00345

994.17

1.0059

suh^m^ ■ calibration of flasks of capacity other than I litre, the corresponding multiple or

above values is taken.

apparent weight in grams in air against brass weights, density
per million nrodern basis for density of weights of 8.0 g cm “ ^ will result in a difference off. 1 parts
More e! Vi * obviously not affect the table to the significant figures quoted.

Volumetric efasswa'^'f (tables for Use in the Calibration of

in, 16 QUANTITATIVE INORGANIC ANALYSIS

mark is reached. The funnel is then carefully removed, taking care not to wet the
neck of the flask above the mark, and then, using a dropping tube, water is added
dropwisc until the meniscus stands on the graduation mark. The stopper is
replaced, the flask reweighed, and the temperature of the water noted.

The true volume of the wafer filling the flask to the graduation mark can be
calculated with the aid of Table III. 4. The values in the table have been obtained
by making allowance for (o) the difference in volume of the glass vessel at the
calibration temperature and at 20 'C. (M the density of water at the temperature
of the calibration, and (c) the effect of buoyancy of the air upon the water and the
brass weights. The figures apply to an atmospheric pressure of 760 mm of
mcrcur)’ and a relative humidity of the air of 50 percent ; the usual deviation from
these figures will affect the buoyancy correction (compare .Section III, 11) only
slightly and can be neglected for most purposes.

Ill, 16. PlPETfliS. Pipettes arc of two kinds: fi) those which have one mark
and (klivcr a .small, constant volume of liquid under certain specified conditions
(transfer pipettes); (ii) those in which the steins arc graduated and arc employed
to deliver various small volumes at discretion (grntlualcd or measuring pipettes).
The transfer pipette consists of a cylindrical bulb joined at both ends to narrower
tubing: a ctilibration mark is etched around the upper (suction) tube, while the
lower (delivery) tube is drawn out to a fine tip. The graduated or measuring
pipette is usually intended for the delivery of pre-determined variable volumes of
liquid : it does not find wide use in accurate work for which a burette is generally
preferred. T ransfer pipettes are constructed with capacities of 1,2.5.10, 20. 25, 50
and IQOcm^ ; those of 10, 25 and 50em^ capacity arc most frequently employed
in macro work. They should conform to BS 1 583 and should carry a colour code
ring at the suction end to identify the capacity (BS 3096): as a safety measure an
additional bulb is often incorporated above the graduation mark. They may be
fabricated from lime-soda or Pyre\ glass, and some high-grade pipettes arc
manufactured in Core.s glass (Coming Glass Works, USA). This is glass which
has been subjected to an ion c.xchangc process which strengthens the glass and
also leads to greater surface hardness, thus giving a product which is resistant to
scratcliing and chipping. Pipettes are available to Class A and Class B
specifications: for the latter Grade typical tolerance values tire;

Pipette capacity 5 10 25 50 100 cm-'

Tolerance O.OI 0.04 0.06 0.08 0.12 cm-'

whilst for Class A, tlie tolerances arc approximately halved.

In using such pipettes, they arc first rinsed with the liquid,
then filled by suction to about 1-2 cm above the mark, and the
upper end of the pipette is closed with the tip of tlic dr>’ index
finger (Fig. Ill, 8); any adhering liquid is w-iped from the
outside of the lower stem. The liquid is allow'cd to run out
.slowly by slightly rcUi.xing the pressure of the finger and by
carefully rotating the pipette until the bottom of the meniscus
just reaches the graduation mark; the pipette must be held
vertically so that the mark is at the same level as the eye. Any
drops adhering to the tip are removed by stroking against a

Fig. Ill, 8

COMMON APPARATUS AND BASIC TECHNIQUES HI, 16

glass surface. The liquid is then allowed to run into the receiving vessel, the tip of
the pipette touching the wall of the vessel. When the continous discharge has
ceased, the jet is held in contact with the side of the vessel for 1 5 seconds (draining
time). At the end of the draining time, the tip of the pipette is removed from
contact with the wall of the receptacle; the liquid remaining in the jet of the
pipette must not be removed either by blowing or by other means.

A pipette will not deliver constant volumes of liquid if discharged too rapidly.
The orifice must be of such size that the time of outflow is about 20 seconds for a
10-cm^ pipette, 30 seconds for a 25-cm^ pipette, and 35 seconds for a 50-cm^
pipette.

Various devices are available for handling corrosive or toxic liquids with
transfer pipettes. Some attachments (e.g. the Griffin pipette filler) consist of a
rubber or plastic bulb with glass ball valves operated between finger and thumb :
these control the entry and expulsion of air to and from the bulb, and thus the
flow of liquid into and out of the pipette. In other devices, a piston-control is
attached to the suction end of the pipette. With the ‘Exelo’ safety pipette, the
suction end of the pipette fits snugly into a hollow barrel with an air vent at the
top. With the barrel pushed right down, the tip of the pipette is placed into the
liquid, the vent closed by the fore-finger, and by pulling the barrel slowly
upwards, liquid is sucked into the pipette until it is above the graduation mark ;
the pipette is then controlled by finger pressure on the air vent and operated as a
conventional pipette.

Calibration. Class A pipettes are usually satisfactory for most analytical
purposes and may be purchased with a BST Certificate. When calibration of a
pipette is necessary, the following procedure should be used.

The pipette must first be thoroughly cleaned using one of the cleaning agents
referred to in Section III, 14. If it is necessary to soak the pipette for an extended
period of time, it may be left standing in the cleaning solution contained in a tall
jar : a chromatography jar or a tall measuring cylinder are suitable. Alternatively,
attach a short piece of rubber tubing and a pinch clip to the upper end of the
pipette, and after filling completely with the cleaning solution, close the pinch clip
and clamp the pipette in a vertical position with the jet dipping into cleaning
so ution contained in a beaker. After this treatment, wash the pipette thoroughly
and finally with distilled water.

he pipette is then filled with distilled water, which has been standing in the
a ance room for at least an hour, to a short distance above the mark. Water is
run out until the meniscus is exactly on the mark, and the out-flow is then
opped. The drop adhering to the jet is removed by bringing the surface of some
^ beaker in contact with the jet, and then removing it without
flasW^ pipette is then allowed to discharge into a clean, weighed stoppered
with th weighing bottle) and held so that the jet of the pipette is in contact
orth ^ vessel (it will be necessary to incline slightly either the pipette

ceased'^ Pip^tte is allowed to drain for 1 5 seconds after the outflow has

drain' ' being in contact with the side of the vessel. At the end of the

pipettp^tl,™^ receiving vessel is removed from contact with the tip of the

ensurin' removing any drop adhering to the outside of the pipette and

determf ^rop remaining in the end is always of the same size. To

down ^ jristant at which the outflow ceases, the motion of the water surface
consider^dt K pipette is observed, and the delivery time is

2 0 be complete when the meniscus comes to rest slightly above the end

Ill, 17 quantitative inorganic analysis

of the delivery lube. The draining time of 15 second-': i.s counted from this
moment. The receiving vessel is weighed, and the temperature of the water noted.
The capacity of liie pipette is then calculated with the aid of Table 111,4. At least
two determinations should be made.

Graduated pipettes consist of straight, fairly narrow tubes with no central bulb,
and are also constructed to a standard .specification (BS700): they arc likewise
colour coded in accordance with BS 3996, Three dilfcrent types are available;
Type t delivers a measured volume from a top 7cro to a .sck'cicd graduation
mark ;

Type 2 delivers a measured volume from a selected graduation mark to the jet:
i.c. the zero is at the jet ;

Type 2 calibrated to eonutin a given capacity from the jet to a selected
graduation mark, and thus to remove a selected volume of solution.
Antomatic pipettes. The Dafert pipette (Fig, 111,9) is an automatic version of a
transfer pipette. One side of the two-way tap is connected to a reservoir
containing the solution to be dispensed, and when the tap is in the
appropriate position, solution fills the pipette completely, exccs.s
.solution draining away through the overflow chamber. The pipette
now contains a definite volume of .solution which i.s delivered to the
receiver by appropriate manipulation of the tap. These pipettes,
which arc constructed to confonn to BS 1132. arc available in a
range of sizes from 5- 100 cm’ and are useful in routine work.

Autodispeasers arc also useful for measuring definite volumes of
solutions on a routine basis, solution is forced out of a container by
depressing a syringe plunger: the movement of the plunger and
hence the volume of liquid dispcn.scd. is controlled by means of a
moveable clamp, the plunger is spring loaded, so that when rcle.ised.
Fig. Ill, 9 it returns to its original position and is immediately ready for
operation again.

Tilting pipette's, which arc attached to a reagent bottle, arc only suitable for
delivering approximate volumes of solution.

Ill, 17. BURETTES. Burettes arc long cylindrical tubes of uniform bore
throughout the graduated length, tc-,ninating at the lower end in a glass stop-
cock and a jet; in cheaper varieties, ihc stopcock may be replaced by a rubber
pinch valve incorporating a glass sphere. A diaphragm-type plastic burette tap is
marketed : tliiscan be fitted to an ordinary burette and provides a delicate control
of the outflow of liquid. The merits claimed include: (a) the tap cannot stick,
because the liquid in the burcttccannot come into contact with the threaded part
of the tap : (/>) no lubricant is generally required ; (r) there is no contact between
ground glass surfaces; and {d) burettes and laps can be readily replaced. Burette
taps made of polytctrafluorocthylcnc (PTFE or Teflon) arc also available; these
have the great advantage that no lubricant is required.

It IS sometimes advantageous to employ a burette with an extended jet svhich is
cn twice at nght angles so that the tip of the jet is displaced by some 7. 5-10 cm
rom the body of the burette. Insertion of the lip of the burette into complicated
assemblies of apparatus is thus faciliated, and there is a further advantage, that if
ca e so utions have to be titrated the body of the burette is kept away from the
source o leat Burettes fitted with iwo-wav stopcocks arc useful for attachment

to reservoirs of stock solutions.

COMMON APPARATUS AND BASIC TECHNIQUES HI, 17

As with other graduated glassware, burettes are produced to both Class A and
Class B specifications in accordance with the appropriate standard (BS 846), and
Class A burettes may be purchased with BST Certificate. All Class A and some
Class B burettes have graduation marks which completely encircle the burette;
this is a very important feature for the avoidance of parallax errors in reading the
burette. Typical values for the tolerances permitted for Class A burettes are :

Total capacity 5 10 50 100 cm^

Tolerance 0.02 0.02 0.06 O.lOcm^

for Class B, these values are approximately doubled. In addition to the volume
requirements; limits are also imposed on the length of the graduated part of the
burette and on the drainage time .

When in use, a burette must be firmly supported on a stand, and various types
of burette holders are available for this purpose. The use of an ordinary
laboratory clamp is not recommended : the ideal type of holder permits the
burette to be read without the need of removing it from the stand, and amongst
holders which the authors have found to be particularly satisfactory are the Fisher
burette holder, in both the original and the cheaper students’ version, and the
Gallenkamp burette holder.

Lubricants for glass stopcocks. The object of lubricating the stopcock of a
burette is to prevent sticking or ‘freezing’ and to ensure smoothness in action.
The simplest lubricant is pure Vaseline, but this is rather soft, and, unless used
sparingly, portions of the grease may readily become trapped at the point where
the jet is joined to the barrel of the stopcock, and lead to blocking of the jet.
Various products are available commercially {e.g., Gallenkamp rubber grease)
which are better suited to the lubrication of burette stopcocks. SiUcone-
coniainmg lubricants should be avoided since they tend to ‘creep’ with consequent
contamination of the walls of the burette.

To lubricate the stopcock, the plug is removed from the barrel and two thin
streaks of lubricant are applied to the length of the plug on lines roughly midway
between the ends of the bore of the plug. Upon replacing in the barrel and turning
[he tap a few times, a uniform thin film of grease is distributed round the ground
^ spring or some other form of retainer may be subsequently attached to
e key to lessen the chance of it becoming dislodged when in use.

Reference is again made to the Teflon stopcocks and to the diaphragm type of
Durette tap which do not require lubrication.

The mode of use of a burette is as follows. If necessary, the burette is
a cleaned using one of the cleaning agents described in Section in, 14,

fr ”^scd with distilled water. The plug of the stopcock is removed

sto^ ^ ® y^rfcl, and after wiping the plug and the inside of the barrel dry, the
funn lubricated as described in the preceding paragraph. Using a small
and tV 10 cm^ of the solution to be used are introduced into the burette,

soluti removing the funnel, the burette is tilted and rotated so that the

throunh whole of the internal surface ; the liquid is then discharged

vert/cf;; • ^ After repeating the rinsing process, the burette is clamped

the zer ^ burette holder and then filled with the solution to a little above
stopcock”^^^-i lunnel is removed, and the liquid discharged through the
the iet k lowest point of the liquid meniscus just touches the zero mark ;

'^omplet ensure that all air bubbles have been removed and that it is

® y ull of liquid. To read the position of the meniscus, the eye must be at

ni, 17 QUANTITATIVE INORGANIC ANALYSIS

the same level as the meniscus, in order to avoid errors due to parallax. In the best
type of burette, the graduations arc carried completely round the tube for each
cm-’ and half-way round for the other graduation marks; parallax is thus easily
avoided. To aid the eye in reading the position of the meniscus a piece of white
paper or cardboard, the lower half of which is blackened either by painting with
dull black paint or by pasting a piece of dull black paper upon it, is employed.
When this is placed so that the sharp disiding line is 1-2 mm below' the meniscus,
the bottom of the meniscus appears to be darkened and is sharply outlined
against the white background : the level of the liquid can then be accurately read.
A variety of 'burette readers’ are available from laboratory supply houses, and a
home-made device which is claimed to be particularly cfTcctive has been
described by Woodward and Redman (Ref. 6c). For all ordinary purposes
readings arc made to 0.05 cm’, but for precision work, readings .should be made
to 0.0I-0.02cm’, u.sing a lens to assist the estimation of the subdivisions.

To deliver liquid from a burette into a conical flask or other similar receptacle,
place the fingers of the left hand behind the burette and the thumb in front, and
hold the tap between the thumb and the fore and middle fingers (Fig. Ill, 10). In
this way, there is no tendency to pull the plug out of the barrel of
the stopcock, and the operation is under complete control. Any
drop adhering to the jet after the liquid has been discharged is
removed by bringing the side of the receiving vessel into contact
with the jet. During the delivery of the liquid, the flask may be
gently rotated with the right hand to ensure that the added
liquid is well mixed with any existing content.s of the flask.

Calibration of a burette. If it is necessary to calibrate a
burette, it is essential to establish that it is satisfactory' with
regard to (a) leakage, and (h) delivery time, before undertaking
the actual calibration process. The burette must naturally be
subjected to a thorough cleaning and rinsing procedure, and
then to test for leakage, the plug is removed from the barrel of
the stopcock and both parts of the stopcock arc carefully
cleaned of ait grease: after wetting well with dc-ionised water,
the stopcock IS reassembled. Tire burette is placed in the holder,
filled with distilled (dc-ionised) water, adjusted to the zero
mark, and any drop of water adhering to the jet removed with a piece of filter
paper. The burette is then allow ed to stand for ten minutes, and if the meniscus
has not fallen by more than one half of a scale division, the burette may be
regarded as satisfactory as far as leakage is concerned.

To test the delivery time, again separate ilie components of the stopcock, dry.
grease and reassemble, then fill the burette to tlic zero rnark w'ith distilled water,
and place in the holder. Adjust the position of tlic burette so that the jet comes
inside the neck of a conical flask standing on the base of the burette stand, but
oes not touch the side of the flask. Open tlie stopcock fully, and note the time
a or the meniscus to reach the low'cst graduation mar)c of the burette; this
f. 1 ?^ closely \rith the time marked on the burette, and in any ease, must

tall within the limits laid dowm by BS 846.

(U ^*,*^*’ these two tests, the calibration may Ire proceeded w’ith. Fill
e urette with distilled water which has been aliow'cd to .stand in the balance
room lor at least an hour to acquire room temperature; ideally, this should be as
near to 20 'C as possible. Weigh a clean, dry stoppered flask of about 100cm’

COMMON APPARATUS AND BASIC TECHNIQUES HI, 18/19

capacity, then after adjusting the burette to the zero mark and removing any
drop adhering to the jet, place the flask in position under the jet, open the
stopcock fully and allow water to flow into the flask. As the meniscus approaches
the desired calibration point on the burette, reduce the rate of flow until
eventually it is discharging dropwise, and adjust the meniscus exactly to the
required mark. Do not wait for drainage, but remove any drop adhering to the jet
by touching the neck of the flask against the jet, then re-stopper and reweigh the
flask. Repeat this procedure for each graduation to be tested; for a 50 cm^
burette, this will usually be every 5 cra^. Note the temperature of the water, and
then, using Table III, 4, the volume delivered at each point is calculated from the
weight of water collected. The results are most conveniently used by plotting a
calibration curve for the burette.

m, 18. WEIGHT BURETTES. For work demanding the highest possible
accuracy in transferring various quantities of liquids, weight burettes are
employed. As their name implies, they are weighed before and after a transfer of
liquid. A very useful form is shown diagrammatically in Fig. Ill, 1 l(n). There are
two ground-glass caps, the lower one is closed, whilst the upper one is provided
with a capillary opening; the loss by evaporation is accordingly negligible. For
hygroscopic liquids, a small ground-glass cap is fitted to the top of the capillary
tube. The burette is roughly graduated in 5-cm^ intervals. The titre thus obtained
is in terms of weight loss of the burette, and for this reason the titrants are'
prepared on a weight/weight basis rather than a weight/volume basis. The errors
associated with the use of a volumetric burette, such as those of drainage,
reading, and change in temperature, are obviated, and weight burettes are
especially useful when dealing with non-aqueous solutions or with viscous
liquids. The advantages of weight titrations are discussed in Ref. 7.

. An alternative form of weight burette due to Redman
(Ref. 6d) consists of a glass bulb, flattened on one side so
that it will stand on a balance pan. Above the flattened
side is the stopcock-controlled ciischarge jet, and a filling
orifice which is closed with a glass stopper. The stopper
and short neck into which it fits are pierced with holes, by
alignment of which air can be admitted, thus permitting
discharge of the contents of the burette through the
delivery jet.

The Lunge-Rey pipette is shown in Fig. Ill, 11(6).
There is a small central bulb (5-1 0 cm^ capacity) closed by
two stopcocks 1 and 2 ; the pipette 3 below the stopcock
has a capacity of about 2 cm^, and is fitted with a ground-
on test-tube 4 . This pipette is of particular value for the
weighing out of corrosive and fuming liquids.

cont'roll H K ^ ®DRETTES, In piston burettes, the delivery of the liquid
uniform *®ovement of a tightly fitting plunger within a graduated tube of

'irive and°^^ ^ particularly useful when the piston is coupled to a motor
in this form serve as the basis of automatic titrators such as the

%in,ii
Dl. IQ

in, 20/21 QUANTITATIVE INORGANIC ANALYSIS

inslrumcnts supplied inter alia hy Mcttlcr Ltd. Metrohm Ltd, Radiometer Ltd.
These instruments can provide automatic plotting of titration curves, and
provision is made for a variable rate of delivery as the end-point is approached so
that there is no danger of overshooting the end-point.

Ill, 20. GRADUATKD (MEASURING) CYLINDERS. These arc graduated
vessels available in capacities from 2 to 2000 cm^ Since the area of the surface of
the liquid is much greater than in a graduated flask, the accuracy is not very high.
Graduated cylinders cannot therefore he employed for work demanding even a
moderate degree of accuracy. They arc, however, useful where only rough
measurements arc required.

WATER FOR LABORATORY USE

111,21. PURIFIED W.ATER. From the earliest days of quantitative chemical
measurements it has been recognised that some form of purification is required
for water which is to be employed in analytical operations, and with increasingly
lower limits of detection being attained in insinimcntal methods of analysis,
correspondingly higher standards of purity are imposed upon the water used for
preparing solutions. Standards have now been laid down for water to be used in
laboratories (Ref. 8). which prescribe limits for non-volatile residue, for residue
remaining after ignition, for pH and for conductivity. The British Standard 3978
give.s the limit for non-volatilc residue as .‘>mgl' '. for residue after ignition as
2mgr for pH, 5.0 -7.5, and for conductivity, 10 megohm'’ per centimetre.

For many years the sole method of purification available was by distillation,
and distilled water was universally employed for laboratory purposes. The
modem water-still is usually made of glass, is heated electrically, and provision is
made for interrupting the current in the event of failure of the cooling water, or of
the boiler-feed supply; the current is also cut olTwhen the receiver is full.

Pure water can also be obtained by allowing tap water to percolate through a
mixture of ion-exchange resins; a strong acid resin which will remove cations
from the water and replace them by hydrogen ions, and a strong base resin (OH“
form) which will remove anions. A number of units arc commercially available
(Perrnutil, Elgaslat, etc.) for the production of dc-ionised water, and the usual
practice is to monitor the quality of the product by means of a conductivity
meter. The resins arc usually supplied in an interchangeable cartridge, so that
maintenance is reduced to a minimum. A mixed-bed ion-exchange column fed
with distilled water is capable of producing water with the verv' low conductivity
of about 0.2 X 10 ’’ohm' ’em"', but in spite of this very low conductivity, the
water may contain traces of organic impurities which can be detected by means of
a spectrofiuorimctcr. For most purposes however the traces of organic material
present in de-ionised water can be ignored, and it may be used in most situations
where distilled water is acceptable.

An alternative method of purifying water is by reverse osmosis. Under normal
conditions, if an aqueous solution is separated by a semi-permeable membrane
rom pure water, osmosis will lead to water entering the solution to dilute it. If
however, sufficient pressure is applied to the solution, i.c. a pressure in excess of

COMMON APPARATUS AND BASIC TECHNIQUES IE, 22

its osmotic pressure, then water will flow through the membrane from the
solution; the process of reverse osmosis is taking place.

This principle has been adapted in the Milli-Q3 system of the Millipore
Corporation (Bedford, Massachusetts) as a method of purifying tap water. The
tap water, at a pressure of 3-5 atmospheres, is passed through a tube containing
the semi-permeable membrane. The permeate which is collected usually still
contains traces of inorganic material and is therefore not suitable for operations
requiring very pure water, but it will serve for many laboratory purposes, and is
very suitable for further purification by ion-exchange treatment. In the Milli-Q2
system, water produced by reverse osmosis is passed first through a bed of
activated charcoal which removes organic contaminants, and is then passed
through a mixed bed ion-exchange column ; the resultant effluent will then meet
the most stringent requirements.

%in,i2

in, 22. WASH BOTTLES. A wash bottle is a flat-bottomed flask fitted up to
deliver a fine stream of distilled water or other liquid for use in the transfer and
washing of precipitates. A convenient size is a 500-750 cm^ flask of Pyrex or
other resistance ^ass; it should be fitted up as shown in Fig. Ill, 12. A rubber
bung is used, and the glass, tubes above the bung should be in the same straight
line and lie in the same plane. The jet should deliver a fine
stream of water; a suitable diameter of the orifice is 1 mm.
All glass tubing must be rounded in the Bunsen flame
after cutting. Thick string, foam rubber, thin sheet cork,
or other insulating material, held in place by copper wire,
should be wrapped round the neck of the flask in order to
protect the hand when hot water is used. Asbestos paper is
best applied wet and allowed to dry overnight; there is
sufficient adhesive material in the paper to make it cling
tightly. In order to protect the mouth from scalding by the
back rush of steam through the mouth-piece when the
blowing is stopped, it is convenient to use a three-holed
... _ rubber stopper; a short piece of glass tubing open at both

is the third hole. The thumb is kept over this tube whilst the water

relea'^H ^ ^ removed immediately before the mouth pressure is

purer ■ '''^sh bottles, fitted with ground-glass Joints, can be

AdTu. should be used with orgaruc solvents that attack rubber,
with a { hottle is available commercially and is inexpensive. It is fitted

be held ^ plastic jet, and has flexible sides. The bottle can
control/ hi ■ ’ ^PP^^^^bion of slight pressure by squeezing gives an easily

liquids A water. It is more or less unbreakable and is inert to many wash
Polvth wash bottle should be used only for cool liquids,

water bottles are sometimes charged with wash liquids other than

solutions drawn to the fact that the components of some wash

bottle whe polythene and may be released into the space in the

cemove thp° ^ fillings and rinsings may be required to

reserve it bhe bottle. It is safer to label the wash bottle and to

solution satu ^ wash liquid. Such wash solutions include a weakly acid
bromine hydrogen sulphide, dilute aqueous ammonia, saturated

Or, and dilute nitric acid.

Ill, 23 QUANTITATIVE INORGANIC ANALYSIS

GENERAL APPARATUS

HI, 23. GLASSWARE. CERAMICS, PLASTIC WARE. In the following
sections, a brief account of general laborator>- apparatus rcieviint to quantitative
analysis will be given. The commonest materials of construction of such
apparatus arc glass, porcelain, fused silica, and various pia.stics; the merits and
disadvantages of these arc considered below.

Glassware. In order to avoid the introduction of impurities during
analysis, apparatus of resistance glass should be employed. For most purposes
Pyrex glass (a borosilicatc glass) is to be preferred. Resistance glass is very
slightly an'ccicd by all solutions, but. in general, attack by acid solutions is less
than that by pure water or by alkaline solutions: for this reason the latter should
be acidified whenever possible, if they must be kept in glass for any length of time.
Attention should also be given to watch, clock, and cover glasses; these should
also be of resistance glass. As a rule, glassware should not be heated with a naked
(lame; a wire gauze, preferably with an asbestos centre, should be interposed
between the flame and the glass vessel.

For special purposes. Corning Vycor glass (96 per cent silica) may be used. It
has great resistance to heat and equally great resistance to thermal shock, and is
unusually stable to acids (except hydrofluoric acid ). water, and various solutions.

The most satisfactory beakers for genera! use are those provided with a spout.
The advantages of this form are; {«) convenience of pouring. (/>) the spout forms
a convenient place at which a stirring rod may protrude from a covered beaker,
and (c) the spout forms an outlet for ste.am or escaping gas when the beaker is
covered witii an ordinary clock gla.ss. The size of a beaker must be selected with
due regard to the volume of the liquid whicli it is to contain. The most useful sizes
arc from 250 to 600 cm^.

Conical (or Erlcnmcyer's) flasks of 200-500-cm'’ capacity find many appli-
cations, fore.xample, in titrations.

Funnels should enclose an angle of 60 . The most useful sizes for quantitative
analysis arc those with diameters of 5.5, 7 and 9cm. The stem should have an
intenial diameter of about 4 mm and should not be more than I5cm long. For
filling burettes and transferring solids to graduated flasks, a short-stem, widc-
ncckcd funnel is useful.

Porcelain apparatas. Porcelain is generally employed for operations in
which hot liquids arc to remain in contact with the vessel for prolonged periods.
It is usually considered to be more resistant to solutions, particularly alkaline
solutions, than glass, although this will depend primarily upon the quality of the
glaze. Shallow porcelain ba.sins with lips arc employed for evaporations.
Casseroles arc lipped, flat-bottomed porcelain dishes provided with handles;
they are more convenient to use Ilian dishes.

Porcelain crucible arc very frequently utilised for igniting precipitates and
eating small quantities of .solids because of their cheapness and their ability to
wit stand high temperatures without appreciable cliange. Some reactions, such
as lusion with sodium carbonate or other alkaline substances, and also
evaporations with hydrofluoric acid cannot be carried out in porcelain crucibles
owing to the resultant chemical attack. A slight attack of the porcelain also taka
place with pyrosulphate fusions.

Fuscd-silica apparatus. Two varieties of silica apparatus arc available
commercially, the translucent and tiie transparent grades. The former is much

COMMON APPARATUS AND BASIC TECHNIQUES HI, 24

cheaper and can usually be employed instead of the transparent variety. The
advantages of silica ware are: {a) its great resistance to heat shock because of its
very small coefificient of expansion, {b) it is not attacked by acids at a high
temperature, except by hydrofluoric acid and phosphoric acid, and (c) it is more
resistant to pyrosulphate fusions than is porcelain. The clrief disadvantages of
silica are: (a) it is attacked by alkaline solutions and particularly by fused alkalis
and carbonates, {b) it is more brittle than ordinary glass, and (c) it requires a
much longer time for heating and cooling than does, say, platinum apparatus.

Coming Vycor apparatus (96 per cent silica glass) possesses most of the merits
effused silica and is transparent. The smallest Vycor crucible has a capacity of
30cm\ but pure silica crucibles as small as 5 cm^ are produced.

Plastic apparatus. Plastic materials are widely used for a variety of items of
common laboratory equipment such as aspirators, beakers, bottles, Buchner
funnels and flasks, centrifuge tubes, conical flasks, filter crucibles, filter funnels,
measuring cylinders, scoops, spatulas, stoppers, tubing, weighing bottles, etc. ;
such products are often cheaper than their glass counterparts, and are frequently
less fragile. Although inert towards many chemicals, there are some limitations
on the use of plastic apparatus, not the least of which is the generally rather low
maximum temperature to which it may be exposed : salient properties of the
commonly used plastic materials are summarised in Table III, 5.

Table HI, 5 Plastics used for laboratory apparatus

Material

Appearance'*’ Highest Chemical reagents'’*’

Attacking

Polythene (L.D.)
Polythene (H.D.)
Polypropylene
TPX(Polymethylpente
Polystyrene
PIPE (Teflon)
Polycarbonate
PVC

(Polyvinylchloride)

Nylon

temperature

Acids

Alkalis

organic

solvents^“^

Weak

Strong

Weak

Strong

80-90

1,2

TL-0

100-110

T-TL

120-130

170-180

1,2

Most

250-300

120-130

Most

T-0

50-70

2, 3,4

TL-O

120

(b) A = transparent; TL = translucent.

^ ' resistant; R = resistant; R* = generally resistant but attacked by

(^)™tongnuxtures;V = veryiesistant.

resistan™'^*™™^’^ ~ ehlorohydrocarbons ; 3 = ketones; 4 = cyclic ethers ; V = very

bckine drawn to the extremely inert character of Teflon, which is so
which si' pressure digestion vessels in
coucpntr ^re decomposed by heating with hydrofluoric acid, or with

concentrated nitric acid (see Section HI, 35).

purposes ar ■'^PP-^ATUS. Crucibles and basins required for special
holds pride fabricated from various metals,, amongst which platinum

0 place by virtue of its general resistance to chemical attack.

COMMON APPARATUS AND BASIC TECHNIQUES HI, 24

Bunsen flame Incorrect
. position

□

Correct

position

Fig.ra,13

Correct
position
Meker or
Fisher flame

3 . Ignition of (a) barium sulphate and sulphates of metals which are not
readily reducible, (b) the carbonates, oxalates, etc., of calcium, barium and
strontium, and (c) oxides which are not readily reducible, e.g., CaO, SrO, AI2O3,
Mn304, Ti02, Zr02, Th02, M0O3, and WO3. (BaO, or compounds
which yield BaO on heating, attack platinum.)

Platinum is attacked under the following conditions, and such operations must
not be conducted in platinum vessels :

I. Heating with the following liquids : {a) aqua regia, {b) hydrochloric acid and
0x1 ising agents, (c) liquid mixtures which evolve bromine or iodine, and {d)
concentrated phosphoric acid (slight, but appreciable action after prolonged

following solids, their fusions, or vapours : {a) oxides,
fpmwirn j nitrates, nitrites, sulphides, cyanides, hexacyano-

/ptwJ . hexacyanoferrate(II) of the alkali and alkaline-earth metals
silve/r”™^^ hydroxides of calcium and strontium); {b) molten lead,
UDon'rerf^^r'^’ ^i"> or gold, or mixtures which form these metals

form tb ^ Pi^osphorus, arsenic, antimony, or silicon, or mixtures which

silicatec^^^tn ^P°” reduction, particularly phosphates, arsenates, and

and telbr reducing agents ; {d) sulphur (slight action), selenium,

which dpp'*'”’ ^ halides (including iron(III) chloride), especially those

carbonatp°™^u^ ^ sulphides or mixtures containing sulphur and a

heatine in y^roxide ; and (g) substances of unknown composition ; {h)
wherpbvu,^” atmosphere containing chlorine, sulphur dioxide, or ammonia.
Solid is rendered porous.

teraperatuL°”’ produced, presents a hazard. It may be burnt off at low

awerbeignit’ access to air, without harm to the crucible, but it should
manner- str ^ Precipitates in filter, paper should be treated in a similar

removed Ash'^ ^SJ^ition is only permissible after all the carbon has been
eonductedin presence of carbonaceous matter should not be

"'illattarlf ft, ^ ? ^fmum crucible, since metallic elements which may be present
tnj conditions,

dishes , of .platinum ware. All platinum apparatus (cru-

Platinutn cruc>>? k clean, polished, and in proper shape. If, say, a

•he crucible th ^ stained, a little sodium carbonate should be fused in

residual solid ° poured out on to a dry stone or iron slab, the

r^oncentrated hv(t^^ u water, and the vessel then digested with

y rochloric acid ; this treatment may be repeated, if necessary. If

ni. 24 QUANTJTATIVfi INORGANIC ANAI.VSIS

hnsa very hish melting poliV^ 77 "xTbrnT^ electrodes- ii

use. nnd is therefore ahiiv.s ha dcLi ‘ 'oo -soft for S

or gold. These allojs are' .slie^^^ ‘Quantities of rhodium. iS,m

retain most of the advantaceous'propcrlie ‘“..'^"’^’‘^''‘^urcs above liOO'C but

ac d ((he cxccption.s are dealt with belouo M Mroiiuori-

c.Mremely small adsorption of wit-'r vm ' jrondnctiviiv ofhc.it ard

area of p.,00cnr mid. hclnl"^

uppreciable if the crudblel^^;;? ^ 'o volatility nuiv b"

magnitude of this loss w,II be cv i, c„,l: f’v '"‘''"m consent 'The

approx, mate loss in weight Ixf ' hi which

temperature indicated; ^ ‘ m mg/10t.)cm-/|iour at thJ

|Siaip5SS5S?=:

causinfi , ■ ^«ult ,n the disi„ ' Ir ‘'’^''’ut-Tcone ofa gas flame

Platiinf, ' owing, pro N

It must be apprec-. ’ '^""”•''*0" ofu carbide of

so rapid that 1111 / ‘^’‘■'‘^'■“oible: inan on-n a ps flame there is

covered criieihb appreciable Thn

advisable, therefor/'^ ^ similarlv p-inlv

crucible in a shnlh^'"' '‘'‘'^"'‘'o^ of irorcomn ^^T'^ ‘o <I>c sulphide. If is.

PJadnum apparauis ^ b oQ-air '

met'abolamj ?’ m •''“‘^'umcrrlonm r''’^ fom

m Ibe last case and (c^oTlkain" Q^’orax and lithium

sulphate). 700 which i d to 'Vf «8‘^nsulphalcs (slight attack

2- Evaporations witi by the addition of ammonium

<'? -id in ,hn

^I'ght attack may occim) " ooncenirated

COMMON APPARATUS AND BASIC TECHNIQUES m, 24

Fig. m, 13

Correct

position

Correct
position
Meker or .
Fisher flame

3 . Ignition of (a) barium sulphate and sulphates of metals which are not
readily reducible, (^t) , the carbonates, oxalates, etc., of calcium, barium and
strontium, and (c) oxides which are not readily reducible, e.g., CaO, SrO, AI2O3,
Mn304, Ti02, Zr02, ThOaj M0O3, and WO3. (BaO, or compounds
which yield BaO on heating, attack platinum.)

Platinum is attacked under the following conditions, and such operations must
not be conducted in platinum vessels :

1 . Heating with the following liquids : {q) aqua regia, {b) hydrochloric acid and
oa ising agents, (c) liquid mixtures which evolve bromine or iodine, and {d)
heatingr^^^ appreciable action after prolonged

following solids, their fusions, or vapours : (a) oxides,
fpm/mn ayaroxides, nitrates, nitrites, sulphides, cyanides, hexacyano-
fpyppnt hexacyanoferrate(II) of the alkali and alkaline-earth metals
silver hydroxides of calcium and strontium); {b) molten lead,

UDon’reH^^F*^’ or gold, or mixtures which form these metals

form th ^ Phosphorus, arsenic, antimony, or silicon, or mixtures which

silicates^^^tn ^P°” reduction, particularly phosphates, arsenates, and
and telhu ^ of reducing agents ; {d) sulphur (slight action), selenium,

which ^ halides (including iron(III) chloride), especially those

^ sulphides or mixtures containing sulphur and a
f'ealine in hydroxide; and (g) substances of unknovra composition: {h)
wherpbv ti ^toosphere containing chlorine, sulphur dioxide, or ammonia.
Solid is rendered porous.

temperaturp°°’ produced, presents a hazard. It may be burnt off at low

Dover be irniit’ access to air,' without harm to the crucible, but it should

manner- str ^ Precipitates in filter, paper should be treated in a similar

mmoved AsV^ ^Spition is only permissible after all the carbon has been
conducted in presence of carbonaceous matter should not be

'vill attaph ft, ^ ? atinum crucible, since metallic elements wliich may be present
' ™der reducing conditio

cibles, dishes of -platinum ware. All platinum apparatus (cru-

Platinum crucfw k polished, and in proper shape. If, say, a

the crucible th ^ stained, a little sodium carbonate should be fused in

tcsidual solid 5- ®°lid poured out- on to a dry stone or iron slab, the
concentrated h . °rit with water, and the vessel then digested with
y rochloric acid : this treatment may be repeated, if necessary. If

Ill, 25 QUANTITATIVE INORGANIC ANALYSIS

fusion with sodium carbonate is without cfTcct, potassium hydrogen sulphate
mav be substituted; a slight attack of the platinum will occur. Disodium
tetraborate may also be used. In some eases, the use of hydrofluoric acid or
potassium hydrogen fluoride may be necessary. Iron stains may be removed by
heating the covered crucible with a gram or two of A.R. ammonium chloride and
applying the full heat of a burner for2-.l minutes.

All platinum vessels must be handled with care to prevent deformation and
denting. Platinum crucibles must on no account be squeered with the object of
loosening the solidified cake after a fusion. Box-wood formers can be purchased
for crucibles and dishes ; these arc invaluable for rc-shaping dented or deformed
platinum ware.

Platinum-clad .stainless steel laboratory ware is available for the evaporation of
solutions of corrosive chemicals. These vessels have all the corrosion-resistance
properties of platinum up to about 550 C. The main features arc: (i) much lower
cost than similar apparatus of platinum; (ii) the overall thickness is about four
times that of similar all-p!atiniim apparatus, thus leading to greater mechanical
strength; and (iii) less susceptible to damage by handling with tongs, etc.

Silver apparalirs. The chief uses of siher crucibles and dishes in the
laboratory arc in the evaporation of alkaline solutions and for fusions with
caustic alkalis ; in the latter ease, the silver is slightly attacked. Gold vessels (m.p.
J050"C) arc more rc.sistanl than silver to fused alkalis. Silver melts at 960 *^C. and
care should therefore Ik taken when it is heated over a bare flante.

Nickel ware. Crucibles and dishes of nickel arc employed for fusions with
alkalis and with sodium pcro.xidc. In the peroxide fusion a little nickel is
introduced, but this is usually not objectionable. No metal entirely withstands
the action of fused sodium peroxide. Nickel oxidises in air, hence nickel
apparatus cannot be used for operations involving weighing.

Iron ware, iron crucibles may be substituted for those of nickel in sodium
peroxide fusions. They arc not so durable, but arc much cheaper.

Stainless-steel ware. Beakers, crucibles, dishes, funnels, etc., of stainless
steel arc available commercially and have obvious uses in the laboratory. They
will not rust, are tough, strong, and highly resistant to denting and .scratching.

ITI, 25. HEATING APP.ARATUS. Various methods of heating are required
in the analytical laboratory ranging from gas burners, electric hot plates and
ovens to muffle furnaces.

Burners. The ordinary Bunsen burner is widely employed for the attain-
ment of moderately high temperatures. The maximum temperature is attained by
adjusting the regulator so as to admit rather more air than is required to produce
a non-iuminous flame; too much air gives a noisy flame, which is unsuitable.

Owing to the differing combustion characteristics and calorific values of the
various gaseous fuels which arc commonly available (town gas, natural gas,
liquefied petroleum (bottled) gas), slight variations in dimensions, including jet
size and aeration controls, arc necessary; for maximum cfflcicncy it is e,sscntia!
that, unless the burner is of the ‘All Gases’ type which can be adjusted, the burner
should be the one intended for the available gas supply.

An improvement in design has been cfl'ccted in burners in which both the gas
and air supply can be regulated. The flow of gas is controlled at the base of the
burner by means of a screw which operates a needle valve; the supply of air is
regulated by screwing the tube of the burner up or down and thus allowing more

COMMON APPARATUS AND BASIC TECHNIQUES HI, 25

or less air to enter through the holes at the base. The Pittsburgh universal burner
(sometimes termed a Tirril burner) is of this type. A temperature of 1 050-1 1 50 °C
inacovered platinum crucible or 600-700 °C in a covered porcelain crucible can
kattained with these burners.

WithaMeker burner a temperature of 1 100-1200 °C is said to be reached in a
covered platinum crucible and 800-900 °C in a covered porcelain crucible. The
volume of air passing through a fully aerated ordinary Bunsen burner is about
2.5 times the volume of the gas (town gas); this is not sufficient for the complete
combustion of the gas, but if attempts are made to increase the aeration, the
flame ‘strikes back’ and burns at the bottom of the tube. In the Meker burner the
holes for the admission of air are large enough to pass sufficient air for the
complete combustion of the gas, and the tube is narrowest near the base and
uidens out near the.top, thus resulting in a more perfect mixing of the gas with
air; a nickel grid is fitted into the top of the burner in order to prevent the flame
striking back. The gas burns in many small flames, with the top of each inner
reducing cone about 1 mm above the top of the burner. The numerous small
flames combine to give a very hot and highly concentrated flame, which is
oxidising in character except below the tips of the tiny flames ; the maximum
temperature is attained just a little above the small flames, i.e., about 2-3 mm
from the top of the burner. The burner is used for the ignition of precipitates that
require a high temperature for conversion into a weighable form, and also for
some fusions.

The ‘Amal’ burner attempts to combine the chief features of the improved
Bunsen burner and the Meker burner. The flame can be turned down very low
without flashing back, and it also furnishes a very hot flame.

The so-called ‘electric Bunsen burner’ is an electric heating unit designed so that
meheatis directed by radiation and convection into a small volume. In one form,
onzontal radiation from a vertically mounted tubular heating element is
concentrated over a very small area by reflection from a polished, anodised
parabolic reflector of pure aluminium : the glowing element may be fitted with a
quartz sleeve to protect it against spillage. Attachments are available for heating
a™?b r ‘^™''®rsion into a hot plate, and also into a small water bath. In
er form, a replaceable heating element (conically shaped and in a refractory
IS mounted in the top of a cylindrical housing, which forms the
pporting case. The housing is provided with air-circulation holes in the lower
between the heating element and the housing prevents any
theel^ ^^Psrature rise. A variable transformer or an energy regulator controls
include supplied, and hence the temperature. The attractive features

nnsitin ^ ^°(*^®ntrated source of heat, cleanliness, absence of smell, use in any
Pojon. and independence of draughts.

silica (‘red rod’), consisting of a radiant heater encased in a

(cxcent h^H ’ fl The direct heating of most acids and other liquids

passes th^ ’'ofluoric acid and concentrated caustic alkalis). Infrared radiation
“Pheatisr^^ f sheath with little absorption, so that a large proportion

fiolent them^i liquid by radiation. The heater is almost unaffected by

Steam ^ ^Tie to the low coefficient of thermal expansion of the silica,

heating sol, baths. Boiling water or steam baths are employed for

Iheir volum *°f ^ below boiling, for slow evaporation of liquids to reduce

l>2akerin wfo’u of precipitates, etc. The simplest form is a lipped

^ water is boiled, the vessel being supported on the rim. Some of

in, 25 QUANTITATIVE INORGANIC ANALYSIS

those available commercially have a small number of openings on the lop, and
these are fitted with a series of copper or stainlcss-stccI rings: vessels of Various
sires can be heated cither on the surface of the bath or partially immersed. The
bath is partly filled with water and heated by steam or electricity. Electrically
heated water baths should preferably be provided with a constant-level device,
thus eliminating the danger of running dry and consequent overheating; all
should be fitted with a cut-out switch to prevent overheating if the water supply
should fail.

Hot plates. The electrically-heated hot plate, preferably provided with
three controls~‘Low’, ‘Medium' and ‘Migh'—is of great value in the analytical
laboratory. The heating elements and the internal wiring should be totally
enclosed : this protects them from fumes or spilled liquids. Electric hot plates with
‘stepless’ controls arc also marketed; thc.se permit of much greater selection of
surfiicc temperatures to be made. A combined electric hot plate and magnetic
stirrer is described in Section III. 27.

Electric ovens. The most convenient type is an electrically heated,
thermostatically controlled drying oven having a temperature range from room
temperature to about 250-300 °C; the temperature can be controlled to within
±1-2‘’C. They arc used principally for drying precipitates or solids at
comparatively low controlled temperatures, and have virtually superseded the
steam oven.

A recent introduction is the 'Mercury 450' microwave oven (marketed by Baird
and Tatlock Ltd) which is particularly valuable for detennining tltc moisture
content of materials,

Mufllc furnaces. An electrically lieatcd furnace of mufTlc form should be
available in every well-equipped laboratory. The ma.ximum temperature should
bo about 1200 '’C. If possible, a thermo-couple and indicating pyrometer .should
be provided; otherwise the ammeter in the circuit should be calibrated, and a
chart constructed showing ammeter and corresponding temperature readings.
Gas-hcated niufilc furnaces arc marketed; these mav uivc temperatures up to
about 1200T.

Air baUis. For drying solids and precipitates at temperatures up to 250“C
in which acid or other corrosive vapours arc evolved, an electric oven should not
be used. An air bath may be constructed from a cylindrical metal (copper, iron,
or nickel) vessel, wrapped with asbestos cloth (of about i in. thickness) and held
in position by copper wire ligatures. The bottom of the vessel may be pierced with
numerous holes and covered with a circular a.sbeslos board. A silica triangle, the
legs of which arc appropriately bent, is inserted inside the bath for supporting an
evaporating dish, crucible, etc. The whole is heated by a Bunsen flame, which is
shielded from draughts. The insulating layer of air prevents bumping by reducing
the rate .at which heal reaches the contents of the inner dish or crucible. An air
bath of similar construction but with special heat-resistant glass sides may also be
used ; this possesses the obvious advantage of visibility inside the air bath.

Infrared lamps and heaters. Infrared lamps with interna! reflectors are
available commercially and arc valuable for evaporating solutions. The lamp
may be mounted immediately above the liquid to be heated; the evaporation
takes place rapidly, without spattering and also without creeping. Units arc
obtainable which permit the application of heal to both the top and bottom of a
number of crucibles, dishes, etc., at the same time: this assembly can char filter
papers in crucibles quite rapidly, and the filter paper does not catch fire.

COMMON APPARATUS AND BASIC TECHNIQUES ffl, 26

Crucible and beaker tongs. Crucible tongs should be made of solid nickel,
nickel steel, or other rustless ferro-alloy. For handling hot platinum crucibles or
dishes, platinum-tipped tongs must be used.

Beaker tongs are available for handling beakers (Griffin form) of
100-2000 cm^ capacity. They are made of stainless steel and have woven asbestos
mittens. The tongs have stainless-steel jaws covered with asbestos sleeves and
also die-cast aluminium grips. An adjustable screw with locknut limits the span
of the jaws and enables the user to adjust the jaw span to suit the container size.

Ill, 26. DESICCATORS AND DRY BOXES. It is usually necessary to
ensure that substances which have been dried by heating (e.g. in an oven, or by
ignition), are not unduly exposed to the atmosphere, otherwise they will absorb
moisture more or less rapidly. In many cases, storage in the dry atmosphere of a
desiccator, allied to minimum exposure to the atmosphere during subsequent
operations will be sufficient to prevent appreciable absorption of water vapour.
Some substances however are so sensitive to atmospheric moisture that all
handling must be carried. out in a ‘dry box’.

A desiccator is a covered glass container designed for
the storage of objects in a dry atmosphere. A common
form of desiccator (Scheibler pattern) is shown in Fig. Ill,
14, it is usually charged with some drying agent, such
as anhydrous calcium chloride (largely used in elementary
work), silica gel, activated alumina, or anhydrous calcium
sulphate (‘Drierite’). Silica gel, alumina and calcium
sulphate can be obtained which have been impregnated
with a cobalt salt so that they are self-indicating : the colour
changes from blue to pink when the desiccant is ex-
t rnaterial can be regenerated by heating in an electric oven

a 50-1 80 °C (silica gel) ; 200-230 °C (activated alumina) : 230-250 °C (Drierite)
M It IS therefore convenient to place these drying agents in a shallow dish which
B situated at the bottom of the desiccator, and which can be easily removed for

baking as required.

^ e action of desiccants can be considered from two points of view. The
efflount of moisture that remains in a closed space, containing incompletely
pres^'* ■ is related to the vapour pressure of the latter, i.e. the vapour

sure is a measure of the extent to which the desiccant can remove moisture.

, — h) 6 Comparative efficiency of drying agents

CuSO,

(sticks)
”^4(95%)
Silica gel

Residual Drying agent Residual
water per water per

litre of air litre of air

in mg in mg

KOH (sticks)
AI2O3

CaS 04

H2SO4

Mg(C104)2

BaO

0.014

0.005

0.003

0.002

0.0007

0.00002

Ill, 26 QUANTITATlVn INORGANIC ANALYSIS

and therefore its efficiency. A second factor is the weight of water that can be
removed per unit weight of desiccant, i.c., the drying capacity. In general,
substances that form hydrates have higher vapour pressures but also have greater
drying capacities. It must be remembered that a substance cannot be dried by a
desiccant the vapour pressure of which is greater than tiiat of the sub.stancc itself.

The relative efficiencies of various drying agents will be evident from the data
presented in Table III, 6. These were determined by aspirating properly
conditioned air througli U-tubes charged with the desiccants: they arc applic-
able, strictly, to the use of these desiccants in absorption tubes, but the figures
may reasonably be applied as a guide for the selection of desiccants for
desiccators. It would appear from the table that a hygroscopic material such as
ignited alumina should not be allowed to cool in a covered vessel over
‘anhydrous’ calcium chloride; anhydrous magnesium perchlorate or phosphorus
pcnto.xidc is satisfaetorj'.

There is however much controvcr.sy regarding the effectiveness of desiccators.
If the lid is briefly removed from a desiccator then it may takx as long as two
hours to remove the atmospheric moisture thus introduced, and to re-establish
the dry atmosphere; during this period, a hygroscopic substance may actually
gain in weight whilst in tlie dcsicc;itor. It is therefore advisable that any snhstance
which is to be weighed should he kept in a vessel with as tightly fitting a lid as
possible whilst it is in the desiccator.

The problem of the cooling of hot vessels within a dcsiccmtor is also important.
A crucible which has been strongly ignited and immediately transferred to a
desiccator may not have attained room temperature even after one hour. The
situation can be improved by allowing the crucible to cool for a few minutes
before transferring to the desiccator, and then a cooling time of 20-25 minutes ks
usually adequate. The inclusion in the desiamtor of a metal block (c.g.
aluminium), upon which the crucible may be stood, is also helpful in ensuring the
attainment of temperature equilibrium.

If a Schciblcr-typc desiccator is employed as a cooling receptacle for weighing
vessels it may be provided with a porcelain plate on feel, which contains
apertures for crucibles, etc. : the porcelain plate should be wedged into the sides,
if necessary, with cork or some other material. For small desiccators, a silica
triangle, with wire ends suitably bent, may be used. The ground edge of the
desiccator should be lightly coated with white Vaseline or a special grease in
order to make it air tight ; too much grease may permit the lid to .slide.

When a hot object, such as a crucible, is placed in a desiccator, about 5-10
seconds should elapse for the air to become heated and e.xpand before putting the
cover in place. When re-opening, the cover siiould be slid open very gradually in
order to prevent any sudden inrush of air due to the partial vacuum which exists
owing to the cooling ol the expanded gas content of the desiccator, and thus
prevent the precipitate being blown out of the crucible.

A desiccator is frequently also employed for the tliorough drying of solids
for an:ily.sis and for other purposes. Its efficient operation depends upon the
condition of the desiccant; the latter should therefore be renewed at frequent
intervals, particularly if its drying capacity is low. For dealing with large
quantities of solid a vacuum desiccator is advisable.

Convenient types of ‘vacuum’ dc.sicca(ors are illustrated in Fig. Ill, 15. Large
surfaces of the solid can be exposed ; the desiccator may be evacuated, and drying
IS thus much more rapid than in the ordinary Schcibler type. These dc.siccators

COMMON APPARATUS AND BASIC TECHNIQUES HI, 26

are made of heavy glass, plastics, or
even metal, and are designed to with-
stand reduced pressure; nevertheless,
no desiccator should be evacuated
unless it is surrounded by an
adequate guard in the form of a stout
wire cage.

For most purposes the ‘vacuum’
produced by an efficient water pump
(20-30 mm mercury) will suffice; a
guard tube containing desiccant
should be inserted between the pump and the desiccator. The sample to be dried
should be covered with a watch or clock glass, so that no mechanical loss ensues
as a result of the removal or admission of air. Air must be admitted slowly into an
exhausted desiccator; if the substance is very hygroscopic, a drying train should
be attached to the stopcock. In order to maintain a satisfactory vacuum within
the desiccator, the flanges on both the lid and the base must be well lubricated
with Vaseline or other suitable grease. In some desiccators an elastomer ring is
incorporated in a groove in the flange of the lower component of the desiccator;
when the pressure is reduced, the ring is compressed by the lid of the desiccator,
and an airtight seal is produced without the need for any grease. The same
desiccants are used as with an ordinary desiccator.

For the efiicient drying of small quantities of materials, the ‘drying pistol’ (Fig.
Ill, 16) may be used. The substance is placed in a porcelain boat which is inserted
into the heating tube B, and is heated by the vapour of the liquid boiling in A;

boxes), which are especially intended for the manipulation of

Ill, 27 QUANTITATIVE INORGANIC ANALYSIS

materials which arc very sensitive to atmospheric moisture (or to oxygen), consist
of a plastic or metal box provided with a window (of glass or clear plastic) on the
upper side, and sometimes also on the side walls. A pair of rubber or plastic
gloves arc fitted through air-tight seals through the front side of the box, and by
placing the hands and forearms into the gloves, manipulations may be carried
out inride the box. One end of the box is fitted with an air-lock so that apparatus
and materials can be introduced into the box without disturbing the atmosphere
inside, A tray of de.riccant placed inside the box will maintain a dry atmosphere,
but to counter the unavoidable leakages in such a .system, it is advisable to supply
a slow current of dry air to the box; inlet and outlet taps arc provided tocontrol
this operation. If the box is flushed out before use with an inert gas (c.g. nitrogen),
and a slow stream of the gas is maintained whilst the box is in use, materials
which arc sensitive to oxygen can be safety handled. For a detailed discussion of
the construction and uses of glove bo.xcs (sec Ref. 9).

Ill, 27. STIRRING APPARATUS. Many operations involving solutions of
reagents require the thorough mixing of two or more reactants, and apparatus
suitable for this purpose ranges from a simple glass stirring rod to electrically
operated stirrers.

Stirring rods. These arc made from glass rod iI-5 mm in diameter, cut into
suitable lengths. Both ends should be rounded by heating in the Bunsen or
blowpipe flame. The length of the stirring rod should be suitable for the size and
the shape of the vessel for which it is employed, c.g., for use with a Ircakcr
provided with a spout, it should project 3~.5cm beyond the lip when in a resting
position.

A short piece of Teflon or of rubber tubing (or a rubber cap) is fitted
tightly over one end of a stirring rod of convenient size. This is the
so-called policeman; it is used for detaching particles of a precipitate
adhering to the side of a vcssscl which cannot be removed by a stream
of water from a wash bottle; it should not, as a rule, be employed for
stirring, nor should it be .allowed to remain in a solution.

Boiling rod.s. Boiling liqiiid.s and liquids in which a gas, such as
hydrogen sulphide, sulphur dioxide, etc., has to be removed by
: boiling can be prevented from super-heating and 'bumping' by the

use of a boiling rod (Fig, HI. 17). This consists of a piecc*^ of glass
Fig. ni, 17 tubing closed at one end and sealed approximately 1 cm from the
other end; the latter end is immersed in the liquid. When the rod is
removed, the liquid in the open end must be shaken out and the rod rinsed with a
jet of water from a wash bottle. This device .should not be used in solutions which
contain a precipitate.

Stirring may be conveniently effected with the so-called magnetic stirrer. A
rotating field of magnetic force is employed to induce variable-speed stirring
action within either closed or open vc.sscls. The stirring is accomplished with the
aid of a small cylinder of iron sealed in Pyrex glass, polythene, or Teflon, which is
caused to rotate by a rotating magnet. A stirrer, fitted with an electric hot plate, is
depicted in sectional diagram in Fig. 111. 18, A speed control is provided, together
with a dial to indicate the setting. >

The usual type of gla.ss paddle stirrer is akso widely used in conjunction with
electric motor fitted with either a transformer-type, or a solid state
controller. The stirrer may be cither connected directly to the motor shaft o

COMMON APPARATUS AND BASIC TECHNIQUES III, 28

Fig. m, 18

spindle actuated by a gear box which forms an integral part of the motor
nousing; by these means, wide variation in stirrer speed can be achieved.

Under some circumstances, e.g. the dissolution of a sparingly soluble solid, it
may be more advantageous to make use of a mechanical shaker. Various models
are available, ranging from ‘wrist action shakers’ which will accommodate small
OHIO erate size flasks, to those equipped with a comparatively powerful electric
mo or and capable of shaking the contents of large bottles, such as ‘Winchester
quarts’, vigorously.

if- filtration APPARATUS. The simplest apparatus used for
, filter funnel fitted with a filter paper. The funnel should have an

filtraf^ possible, and a long stem (15 cm) to promote rapid

aotirnn" * papers are made in varying grades of porosity, and one
P pnate to the type of material to be filtered must be chosen (see Section

determinations involving the collection and
ruciblp ■ ° ^ it is convenient to be able to collect the precipitate in a

ave C” H weighed directly, and various forms of filter crucible

'hichma purpose. The first of these was the Gooch crucible

orcelain v encountered in porcelain, in silica, and (rarely) in platinum: the
common, and the term ‘Gooch crucible’ is generally

The Gon 1.° "^‘^ible.

umber of. consists of a tall form crucible with the base pierced with a

Jukine a r ’ ^^e covered by a pad of asbestos, produced by

measure: thp"^^ asbestos fibres in water through the crucible under reduced
(Sectir>n^m'^^«s°^®‘^“'^® For preparation of the crucible will be described
>■40). The asbestos employed must be carefully selected and

HI, 28 QUANTITATIVI- INORGANIC ANALYSIS

purified. NOTE The normal recommended precaution.s must be taken when
handling asbcsio.s (Ref. 24).

Sintered gln.ss crucibles arc made of resistance glass and have a porous disc of
sintered ground glass fused into the body of the crucible. The filter disc is made in
varj'ing porosities as indicated by numbers from 0 (the coarsest) to 5 (the finest);
the range of pore diameter for the various grades is as follows;

Porosity 0 I 2 3 4 5

Pore diameter (/nn) 200-250 100-120 40-50 20-30 5-10 1-2

Porosity 3 is suitable for precipitates of moderate particle sire, and porosity 4 for
fine precipitates such as barium sulphate. These crucibles should not be heated
above about 200 'C.

Silica crucible's of similar pattern arc also available, and, although expensive,
have certain advantages in thermal stability.

Filter crucibles uith porous filler base are available in porcelain (porosity 4), in
silica (porosities 1, 2, 3. 4), and in alumina (coarse, medium and fine porosities);,
these have the advantage as compared with sintered crucibles, of being capable of
being heated to much higher temperatures. Nevertheless, the heating must be
gradual otherwise the crucible may cr:ick at the join between porous base and
glazed side.

For filtering large quantities of material, a Ruchner funnel is usually employed ;
alternatively, one of the modified funnels shown diagrammntically in Fig. Ill, 19
may l>c used. I Icre (n) is the ordinary porcelain Buchner funnel ; (/>) is the ‘.sHt sieve'
glass funnel. In both cases, one or, better, two good-quality filler papers arc
placed on the plate; the glass type is preferable since it is transparent and it is easy
to see whether the funnel is j^erfcctly clean, (c) is a Pyrex funnel with a sinlcrcd
glass plate; no filter paper is required so that .strongly acidic and weakly alkaline
solutions can be readily filtered with thi.s funnel. In all cases the funnel of
appropriate size is fitted into a filter flask (</), and the filtration conducted under
the diminished pressure provided by a filter pump or vacuum line.

(It) (c) 1(1)

Fig. Ill, 19

One of the disadvantages of the porcelain Buchner funnel is that, being of one-
piece construction, the filler plate cannot be removed for thorough cleaning and
It IS difficiili to sec whether the whole of the plate is clean on both sides. In a
modern polythene version, the funnel is made in two sections which can be
unscrewed thus permitting inspection of both sides of the plate.

The Hartley funnel, shown in Fig, HI, 20. consists of three detachable parts ; an

COMMON APPARATUS AND BASIC TECHNIQUES III, 29

upper and lower part, both of which have flanges extending
beyond the actual filtering area, and a detachable plate
ground on both sides to fit the flanges on the lower and upper
portions. A filter paper is used which covers the whole
surface of the ground area; creeping of the material to be
filtered underneath the edges of the filter paper is thus
avoided. In use, the filter paper is first wetted with the
appropriate solvent, placed on the filter disc, and then the
flanged ring placed on the paper. Generally the weight of the
ring renders the joint leak-proof, but small clips can be used
to keep the three parts together; it is advisable to apply
suction before the funnel is filled with liquid.

In some circumstances, separation of solid from a liquid is better achieved by
use ofa centrifuge than by filtration, and a small, electrically driven centrifuge is a
useful piece of equipment for an analytical laboratory. It may be employed for
removing the mother liquor from recrystallised salts, for collecting difficultly
filterable precipitates, and for the washing of certain precipitates by decantation.
It is particularly useful when small quantities of solids are involved ; centrifuging,
followed by decantation and re-centrifuging, avoids transference losses and
yields the solid phase in a compact form. Another valuable application is for the
separation of two immiscible phases.

%ni ,20

m, 29. WEIGHING BOTTLES. Most chemicals are weighed by difference
by placing the material inside a stoppered weighing bottle which is then weighed.
The requisite amount of substance is shaken out into a suitable vessel (beaker or
fiask), and the weight of substance taken is determined by reweighing the
weighing bottle. In this way, the substance dispensed receives the minimum
exposure to the atmosphere during the actual weighing process ; a feature of some
importance if the material is hygroscopic.

The most convenient form of weighing bottle is one fitted with an external cap
311 made of glass, polythene or polycarbonate. A weighing bottle with an
mternally fitting stopper is not recommended; there is always the danger that
^ma particles may lodge at the upper end of the bottle and be lost when the

%per is pressed into place.

the substance is unaffected by exposure to the air, it may be weighed on a
2\Y ™ ^ disposable plastic container. The weighing funnel (Fig. Ill,

We * useful, particularly when the solid is to be transferred to a flask : having
•8 ed the solid into the scoop-shaped end which is flattened so that it will stand
on the balance pan," the narrow end is inserted into the
neck of the flask and the solid washed into the flask with
a stream of water from a wash bottle.

Woodward and Redman (Ref. 6e) have described a
specially designed weighing bottle which will accom-
been ig > ^ • modate a small platinum crucible : when a substance has

subsea**” I crucible, the crucible is transferred to the weighing bottle and
If th^'^k '^®*Shed in this. This device obviates the need for a desiccator.
Section Bt weighed is a liquid, a Lunge-Rey pipette (Fig. Ill, 1 1 ;

bottlpfiH j be used. Alternatively, the liquid is placed in a weighing

ed with a cap carrying a dropping tube.

%in,21

Ill, 30 QUANTITATIVE INORGANIC ANALYSIS

Reagents and Standard Solutions

III, 30. REAGENTS. Tlie purest rcagcnis available should be used for
quantitative analysis; the analytical reagent quality (AR) is generally employed.
In Great Britain AR (‘AnalaR’) chemicals from BDH Chemicals Ltd, and from
Hopkin and Williams Ltd conform to the .spcciiicalion.s given in their handbook
‘AnalaR' Standards for Laboratory Chemicals. In the USA the American
Chemical Society committee on Analytical Reagents Itas established standards
for certain reagents, and manufacturers supply reagents which arc labelled
‘Conforms to ACS Specifications'. In addition, certain manufacturers market
chemicals of high purity, and each package of these analy.sed chemicals has a
label giving the manufacturer's limits of certain impurities.

With the increasingly lower limits of detection being achieved in various types
of instrumental analysis, there is an ever growing demand for reagents of
correspondingly improved specification, and some manufacturers arc now
offering a range ofspeciallv purified reagcnl.s. c.g., the BDll Chemicals 'Aristar'
chemicals; Hopkin and Williams PVS (Purified for Volumetric Standardisation)
range.

In some instances, where a reagent of the requisite purity is not available, it
may be advisable to weigh out a suitable portion of the appropriate pare metal
[c.g. thc.lohnson, Matthey ‘Speepure’ range, or the M ARZgradc metals supplied
by the Materials Research Corporation. USA (Materials Research Co, London,
England) ), and to dissolve this in the appropriate acid.

It must be remembered that the label on a bottle is not an infallible guarantee
of the purity of a chemical, for the following reasons;

(o) Some impurities may not have liccn tested for by the manufacturer.

(/)) The reagent may have been contaminated after it.s receipt from the
manufacturers either by the stopper having been left open for some time,
with the con.scqucnt exposure of the contents to the laboratory atmosphere
or by the accidental return of an unused portion of the reagent to the bottle,
(c) In the ease of a solid reagent, it may not be sufficiently dry. This may be due
cither to insufficient drying by the manufacturers or to leakage through the
stoppers during storage, or to both of these causes.

However, ifthcanalytical rcagent.s arc purchased from a manufacturing firm of
repute, and instructions given (n) that no bottle is to be opened for a longer time
than is absolutely ne'ecssary, and (/>) that no reagent is to be returned to the bottle
after it has been removed, the likelihood of any errors arising from some of the
above possible causes is considerably reduced. Liquid reagents should be poured
from the bottle; a pipette should never be inserted into the reagent bottle.
Particular care should be taken to avoid contamination of the .stopper of the
reagent bottle. When a liquid is poued from a bottle, the stopper should never be
placed On the shelf or on the working bench; it may be placed upon a clean
watchglass. and many chemists cultivate the habit of holding the stopper between
the thumb and fingers of one hand. The stopper should be returned to the bottle
immediately after the reagent has been removed, and all reagent bottles should be
kept scrupulously clean, particularly round the neck or mouth ofthe bottle.

If there is any doubt as to the purity ofthe reagents used, they should be tested
by standard methods for the impurities that might cause errors in the
determinations. It may be mentioned that not all chemicals employed in
quantitative analysis are available in the form of analytical reagents; the purest

COMMON APPARATUS AND BASIC TECHNIQUES m, 31

commercially available products should, if necessary, be purified by known
jnethods: see below. The exact mode of drying, if required, will vary with the
reagent; details are given for specific reagents in the text.

m, 31. PURIFICATION OF SUBSTANCES. If a reagent of adequate
purity for a particular determination is not available, then the purest available
product must be purified : this is most commonly done by recrystallisation from
water. A known weight of the solid is dissolved in a volume of water sufficient to
give a saturated or nearly saturated solution at the boiling point : a beaker,
conical flask or porcelain dish may be used. The hot solution is filtered through a
fluted filter paper placed in a short-stemmed funnel, and the filtrate collected in a
beaker : this process will remove insoluble material which is usually present. If the
substance crystallises out in the funnel, it should be filtered through a hot-water
funnel. The clear hot filtrate is cooled rapidly by immersion in a dish of cold
water or in a mixture of ice and water, according to the solubility of the solid ; the
solution is constantly stirred in order to promote the formation of small crystals,
which occlude less mother liquor than larger crystals. The solid is then separated
from the mother liquor by filtration, using one of the Buchner-type funnels
shown in Fig. Ill, 1 9 and Fig. Ill, 20 (Section HI, 28). When all the liquid has been
filtered, the solid is pressed down on the funnel with a wide glass stopper, sucked
as dry as possible, and then washed with small portions of the original solvent to
remove the adhering mother liquor. The recrystallised solid is dried upon clock
glasses at or above the laboratory temperature according to the nature of the
material; care must of course be taken to exclude dust. The dried solid is

preserved in glass-stoppered bottles. It should be noted that unless great care is
taken when the solid is removed from the funnel, there is danger of introducing
fibres from the filter paper, or small particles of glass from the glass filter disc:
scraping of the filter paper or of the filter disc must be avoided.

Some solids are either too soluble, or the solubility does not vary sufficiently
With temperature, in a given solvent for direct crystallisation to be practicable. In
many cases, the solid can be precipitated from, say, a concentrated aqueous
so ution by the addition of a liquid, miscible with water, in which it is less soluble.

anol, in which many inorganic compounds are almost insoluble, is generally
i>se . Care must be taken that the amount of ethanol or other solvent added is not
so arge that the impurities are also precipitated. Potassium hydrogencarbonate
tartrate may be purified by this method,
aublimation. This process is employed to separate volatile substances from
pu^'fi arsenic trioxide and ammonium chloride can be

^ 0 in this way. The substance is placed in a porcelain dish or casserole; the
latter is gently heated with a small flame and the vapour
' condensed upon a cool surface, such as a large inverted
I ^ j glass funnel containing a plug of glass wool at the apex,

Jr or, preferably, a flask containing cold water.

' Pure iodine is required in analysis, and details for its puri-

fication will not be out of place here. Grind together 10 g
iSy iodine with 4 g of potassium iodide (any chlorine or

bromine present will thus be retained as the non-volatile

k j potassium salts) and transfer the mixture to a casserole ;

m, 22 place a flask through which a gentle stream of cold water is

circulating, on the casserole (Fig. Ill, 22). Heat very gently

Ill, 32 QUAMTITATIVE INORGANIC ANALYSIS

until sumcicnt iodine has sublimed on to the bottom of the flask, allow to cool,
and remove the flask with the iodine adhering to it. Pass a rapid stream of ice-cold
water through the flask; this will cause the gla.ss to contract somewhat and the
whole of the crust can then be removed by scraping with a clean glass rod and is
collected on a clock glass. Break up the large pieces, and repeat the sublimation
without the addition of potassium iodide. Remove the second sublimate as
before, and grind the iodine in a glass mortar. Dry in a desiccator containing
calcium chloride; no grease whatever should be exposed on the inside, since
iodine vapour attacks grease forming hydrogen iodide.

Zone refining is a purification technique originally developed for the refinement
of certain metals, and which is applicable to all substances of reasonably low
melting point which arc stable at the melting temperature. In a zone refining
apparatus, the substance to be purified is packed into a column of glass or
stainless steel, which may vary in length from six inches (semimicro apparatus) to
three feet. An electric ring heater which heats a narrow band of the column is
allowed to fall slowly by a motor-controlled drive, from the top to the bottom of
the column. The heater is set to produce a molten zone of material at a
temperature 2-3 C above the melting point of the substtince. which travels
slowly down the tube with the heater. Since impurities normally lower the
melting point of a substance, it follows that the impurities tend to flow down the
column in step with tlic heater, and thus to become concentrated in the lower part
of the tube. The process may be repeated a number of times (the apparattis may
be programmed to reproduce automatically a given number of cycles), until the
required degree of purification ha.s been achieved. The benzoic acid PVS grade
marketed by Hopkin and Williams is purified by this technique (see Ref. 10).

Ill, 32. PREPARATION AND STORAGE OF STANDARD
SOLUTIONS. In any analytical laboratory it is essential to maintain stocks of
solutions of various reagents: some of these will be of accurately known
concentration (i.c. .standard solutions) and correct storage of such solutions is
imperative.

According to BS 4245 solutions should be classified as :

1 . reagent solutions which are of approximate concentration ;

2. standard solutions which have a known concentration of some chemical ;

3. standard reference solutions which have a known concentration of a primary

standard substance (Section X, 6);

4. standard titrimctric solutions which have a known concentration (de-
termined cither by weighing or by standardisation), of a substance other than

a primary standard.

ThelUPAC Commission on Analytical Nomenclature (Ref. 1 1 ), refer to 3 and 4
respectively as Primary' Standard Solutions and Secondary Standard Solutions.

For reagent solulioas as defined above (i.c. 1), it is usually sufllcicnt to weigh
out approximately the amount of material required, using a watch glass or a
plastic weighing container, and then to add this to the required volume of solvent
which has been measured with a measuring cylinder.

To prepare a standard solution the following procedure is followed. A short-
stemmed funnel is inserted into the neck of a graduated flask of the appropriate
size. A suitable amount of the chemical is placed in a weighing bottle which is
weighed, and then the required amount of substance is transferred from the
weighing bottle to the funnel, taking care that no particles arc lost. After the

100

COMMON APPARATUS AND BASIC TECHNIQUES HI, 32

weighing bottle has been re-weighed, the substance in the funnel is washed down
with a stream of the liquid. The funnel is thoroughly washed, inside and out, and
then removed from the flask; the contents of the flask are dissolved, if necessary,
by shaking or swirling the liquid, and then made up to the mark: for the final
adjustment of volume, a dropping tube drawn out to form a very fine jet is
employed.

If a watch glass is employed for weighing out the sample, the contents are
transferred as completely as possible to the funnel, and then a wash bottle is used
to remove the last traces of the substance from the watch glass. If the weighing
scoop (Fig. Ill, 21 ; Section HI, 29) is used, then of course a funnel is not needed
provided that the flask is of such a size that the end of the scoop is an easy fit in the
neck.

If the substance is not readily soluble in water, it is advisable to add the
material from the weighing bottle or the watch glass to a beaker, followed by
distilled water; the beaker and its contents are then heated gently with stirring
until thesolid has dissolved. After allowing the resulting concentrated solution to
cool a little, it is transferred through the short-stemmed funnel to the graduated
flask, the beaker is rinsed thoroughly with several portions of distilled water,
adding these washings to the flask, and then finally the solution is made up to the
mark: it may be necessary to allow the flask to stand for a while before making
the final adjustment to the mark to ensure that the solution is at room
temperature. Under no circumstances may the graduated flask be heated.

In some circumstances it may be considered preferable to prepare the standard
solution by making use of one of the concentrated volumetric solutions supplied
by various manufacturers (BDH Chemicals ‘CVS’ solutions,
Hopkin and Williams ‘Convol’ solutions. May and Baker ‘Volucon’ solutions,
sol require dilution in a graduated flask to produce a standard

Solutions which are comparatively stable and unaffected by exposure to air
®ay be stored in litre, or in ‘Winchester quart’ bottles; for work requiring the

highest accuracy, Pyrex, or other
resistance glass, bottles fitted with
ground-glass stoppers should be
employed, the solvent action of the
solution being thus consider-
ably reduced. It is however neces-
sary to use a rubber bung instead
of a glass stopper for alkaline
solutions, and in many instances
a polythene container (e.g. an
aspirator) may well replace glass
vessels. It should be noted how-
ever that for some solutions as,
for example, iodine and silver
nitrate, glass containers only may
be used, and in both these cases
the bottle should be made of
dark (brown) glass: solutions of
EDTA (Section X, 50) are best
stored in polythene containers.

101

Ill, 32 QUANTITATIVE INORGANIC ANALYSIS

The bottle should be dean and drj*: a little of the stock solution is introduced,
the bottle well rinsed with this solution, drained, and the remainder of the
solution poured in and the bottle immediately stoppered. If the bottle is not dry,
but has recently been thoroughly rinsed with distilled water, it may be rinsed
successively with three small portions of the solution and drained well after each
rinsing; this procedure is, however, less .satisfactory than that employing a clean
and dry vessel. Immediately after the solution has been transferred to the stock
bottle, it should be labelled with: (1) the name of the .solution. (2) its
concentration. (3) the date of preparation, and (4) tlic initials of the person who
prepared the solution, together with any other relevant data. Unless the bottle is
completely filled, internal evaporation and condensation will cause drops of
water to form on the upper part of the inside of the vessel. For this reason, the
bottle must be thoroughly shaken before removing the stopper.

For expressing concentrations of reagents, the molar system is universally
applicable, i.e., the number of moles of solute present in 1 dm^ of solution.
Concentrations may also be e.xprcssed in terms of normality if no ambiguity is
likely to arise (see Section X, 3), and in fact BS2445 (1968) recommends that
concentrations should usually be expressed in normalities.

Solutions liable to be affected by access of air (c.g., alkali hydroxides which
absorb carbon dioxide; iron(ll) and titanium(lll) which arc oxidised) may be
stored in the apparatus shown diagranimatically in Fig. Ill, 23 (n). The burette
has a three-way tap which enables it cither to be filled from the stock bottle or to
be emptied. If such a burette is not available, one wiili a side-tube and stopcock
(6) will serve equally well. The tube T is permanently connected with a source of
hydrogen (c.g., from a Kipp's apparatus) if the solution is o.vidiscd upon
exposure to air, or to a soda-lime or sodium hydroxide- asbestos guard tube, ifit
contains cau-stic alkali. In the latter ease, particularly if soda glass vessels arc
used, the solution may become contaminated with silicates owing to the attack of
the alkali on the glass: it is belter to employ a storage vessel of resistance glass or,
preferably, of polythene.

A more compact apparatus is shown in Fig. HI, 24. A is a large storage bottle
of 1 0- 1 5 litres capacity. 11 is a 50-cm^ burette provided with an automatic filling

device at C (the point of the drawn-out tube is
adjusted to be exactly at the zero mark of the
burette), D is the burette-bottle clamp, E is a
two-holed rubber stopper, F is a ground-glass
tension joint, a rubber tube is conncctal to a
source of hydrogen (for example, a Kipp’s
apparatus) and to the T-joint below L, H is a
Bunsen valve, and J is hydrogen. The burette
is filled by closing tap K and passing hydrogen
through the rubber tube attached to the T-
picce (below tap L) with lap L closed; taps L
and K are opened, and the excess of liquid
allowed to siphon back.

Two other apparatus for the storage of
standard solutions arc .shown in lug. Ill, 25.
Fig. Ul,25(/i)issc!f-c.xpIanatory. Thcsolution
is contained in the storage bottle A, and the
50-cm-^ burette is fitted into this by means of a

102

COMMON APPARATUS AND BASIC TECHNIQUES m, 33

ground-glass joint B. To fill the burette, tap C is opened and the liquid pumped
into the burette by means of the small bellows E. F is a small guard tube ; this is
filled with soda-lime or ‘carbosorb’ when caustic alkali is contained in the storage
bottle. Bottles with a capacity up to 2 litres are provided with standard ground-

Fig. 111,25

g ass joints; large bottles, up to 15 litres capacity, can also be obtained. Fig. Ill,
W portrays a similar apparatus, but with an automatic filling device. The
0 u ion IS pumped into the burette and enters it through a glass tube which
^ capillary exactly at the zero mark; immediately the pressure is
the rtora b above the zero mark is automatically siphoned back into

dispe^ pipette (Fig. Ill, 9; Section III, 16) is a convenient apparatus for
^ volumes of a standard solution, as are also the various liquid
oispensers which are available.

some basic techniques

ra, 33.

Presented OF THE SUBSTANCE FOR ANALYSIS,

inunediater* r Quantity of a material to be analysed, the analyst is
theanalvti with the problem of selecting a representative sample for

tlrat comm^ ruvestigations. It may well be that the material is in such large pieces
Irandling in tt!^ i°u accessary in order to produce a specimen suitable for
c aboratory. These important factors are considered in Chapter V

103

ni, 34/35 QUANTITATIVE INORGANIC ANALYSIS

(Sections V, 2; V, 3), and as explained therein, the material is usually dried at
105-1 lOX before analysis.

in, 34. WEIGHING THE SAMPLE. If necessary refer to Section III, 11
dealing with the operation of a chemical balance, and to Sections III, 29 and III,
26 which are concerned with the use and function of weighing bottles and
desiccators respectively.

The material, prepared ti.s above, is usually transferred to a weighing bottle
which is stoppered and stored in a desiccator. Samples of appropriate size arc
withdrawn from the weighing bottle as required, the bottle being weighed before
and after the withdrawal, so that the weight of substance is obtained by
difference.

Attention isdrawn to the Mctticr vibro-spatula which is a useful adjunct to the
weighing out of powders. The spatula is connected to the electric mains, and the
powder is placed on the blade ofthc spatula. When the current is switched on by
pressing the button 1 , the blade is caused to vibrate and to deposit solid gradually
into the beaker or other container over which it is held: the intensity of the
vibration may be adjusted by means of the knurled heail 2 (Fig, 111, 26).

Fir, III, 26

HI, 35. SOLUTION OF THE SAMPl.E. Whilst many substances can be
dissolved directly in water or in dilute acids, materials such as minerals,
refractories, and alloys must usually be treated with a variety of reagents in order
to discover a suitable solvent : m such cases the preliminary qualitative analysis
will have revealed the best procedure to adopt. Each ease must be considered on
its merits; no attempt at generalisation will therefore be made. We can. however,
discuss the experimental technique of the simple process of solution of a
substance in water or in acids, and also the method of treatment of insoluble
substances.

For a substance which dissolves readily, the sample is weighed out into a
beaker, and the beaker imniedia cly covered with a clock glass ofsuitablc size (its
diameter should not bemorctl.un about 1 cm larger than that of the beaker) with
its convex side facing downwards. The beaker should have a spout in order to
provide an outlet for the escape of steam or gas. The solvent is then added by
pouring it card ully down a gla.ss rod. the lower end of which rests against the wail
ofthc beaker ; the clock glass is displaced somewhat during this process. If a gas is
cv’olved during the addition ofthc solvent (c.g., acids with carbonates, metals,
alloys, etc.), the beaker must be kept covered as far as possible during the
addition. The reagent is then best added by means of a pipette or by means ofa
funnel with a bent stem inserted beneath the clock glass at the spout of the
beaker; loss by spirting or as spray is thus prevented. When the evolution of gas
has ceased and the substance has completely dissolved, the under side ofthc clock
g ass is well rinsed with a stream of water from n wash bottle, care being taken
that the washings fall on to the side of the beaker and not directly into the
solution. If warming is nccc.ssary, it is usually best to carry out the dissolution in a

104

COMMON APPARATUS AND BASIC TECHNIQUES III, 35

conical,(Erlenmeyer) flask with a small funnel in the mouth; loss of liquid by
spirting is thus prevented and the escape of gas is not hindered.

It may often be necessary to reduce the volume of the solution, or sometimes to
evaporate completely to dryness. Wide and shallow vessels are most suitable,
since a large surface is thus exposed and evaporation is thereby accelerated. We
may employ shallow beakers of resistance glass, Pyrex evaporating dishes,
porcelain basins or casseroles, silica or platinum basins ; the material selected will
depend upon the extent of attack of the hot liquid upon it and upon the
constituents being determined in the subsequent analysis. Evaporations should
be carried out on the steam bath or upon a low-temperature hot plate ; slow
evaporation is preferable to vigorous boiling, since the latter may lead to some
mechanical loss in spite of the precautions to be mentioned below. During
evaporations, the vessel must be covered by a Pyrex clock glass of slightly larger
diameter than the vessel, and supported either on a large all-glass triangle or
upon three small U-rods of Pyrex glass hanging over the rim of the container.
Needless to say, at the end of the evaporation the sides of the vessel, the lower side
of the clock glass and the triangle and glass hooks (if employed) should be rinsed
with distilled water into the vessel.

For evaporation at the boiling point either a conical flask with a short Pyrex
funnel in the mouth or a round-bottomed flask inclined at an angle of about 45°
may be employed; in the latter the drops of liquid, etc., thrown up by the
ebullition or by effervescence will be retained by striking the inside of the flask,
while gas and vapour will escape freely.

Substances which are insoluble or only partially soluble in acids are brought
into solution by fusion with the appropriate reagent. The most commonly used
fusion reagents, or fluxes as they are called, are anhydrous sodium carbonate
either alone or, less frequently, mixed with potassium nitrate or sodium
peroxide; potassium, or sodium pyrosulphate; sodium peroxide; sodium or
potassium hydroxide. Of recent years anhydrous lithium metaborate has found
avour as a flux, especially for materials containing silica (Ref. 14); when the
resulting fused mass is dissolved in dilute acids, no separation of silica takes place
as it does when a sodium carbonate melt is similarly treated. Other advantages
aimed for lithium metaborate are :

evolved during the fusion or during the dissolution of the melt,
^ an hence there is no danger of losses due to spitting;

lithium metaborate are usually quicker (15 minutes will often

3 th I ^ performed at a lower temperature than with other fluxes ;

e OSS of platinum from the crucible is less during a lithium metaborate

4 I ^ sodium carbonate fusion ;

vvitu^ elements can be determined directly in the acid solution of the melt
It is cl H tedious separations.

Puritv^'”'^^-^^^*’ circumstances the fusion may be performed in a high

the fliiY 1 1® always satisfactory (Ref. 15). Naturally,

''esselin^ V depend upon the nature of the insoluble substance. The

emploveJ lesion is effected must be carefully chosen ; platinum crucibles are
Phate-ni vT carbonate, lithium metaborate and potassium pyrosul-

silver or' ^ ®ll\£r crucibles for sodium or potassium hydroxide ; nickel, gold,
crucibles ‘^^pibles for sodium carbonate and/or sodium peroxide; nickel
attacked! t"^ ®°clium carbonate and potassium nitrate (platinum is slightly
ocarry out the fusion, a layer ofthe flux is placed at the bottom ofthe

105

Ill, 36 QUANTITATIVK INORGANIC ANALYSIS

crucible, and then ilie intimate mixture of the flux and the finely-divided
substance added; the crucible should be not more than about lialf-full, and
should, generally, be kept covered during the whole process. The crucible is very
gradually heated at first, and the temperature slowly raised to the required
temperature. The final temperature should not be higher than is actually
ncccssaty; any possible further attack of the flux upon the crucible is thus
avoided.' When the fusion, which usually takes 30 -60 minutes, has been
completed, the crucible is grtisped by means of the crucible tongs and gently
rotated and tilted so that the molten material distributes itself around the walls of
the container and solidifies thc.>-c as a thin layer. This procedure greatly facilitates
the subsequent detachment and solution of the fused mass. When cold, the
crucible is placed in a casserole, porcelain dish, platinum basin, or Pyrex beaker
(according to the nature of the flux) and covered with water. Acid is added, if
necessary, and the scsscl is covered with a clock glass, tlie temperature raised to
95-100 "'C, and maintained until solution is achieved.

Many of the substances which require fusion treatment to render them soluble
will in fact dissolve in mineral acids if the digestion with acid is carried out under
pressure, and consequently at higher temperatures tlian those nonnaliy achieved.
Such drastic treatment requires a container cap:iblc of withstanding the requisite
pressure, and also resistant to chemical attack. A satisfactors' solution of this
problem was achieved by Hcmas (Ref. 12). who devised ;i stainless steel pressure
vessel, capacity SOenv’, lilted with a Teflon liner. .Acid digesflon vessels made on
this principle arc now available from Uniscai Decomposition Vessels Ltd. Haifa,
Israel (British agents, S. and J. Juniper and Co. Harlow. Essex), and also from tlie
Parr Instrument Co. U.SA; they may be heated to temperatures of I5()-I80"C,
will withstand pressures of St)-9U atmospheres, and under these conditions
decomposition of refractory matcriaK may beaccompILshcd in 45 minutes. Apart
from tlie saving in time which is achieved, and the fact that the use of expensive
platinum ware is obviated, other advantages of the method arc that no losses can
occur during the treatment, and the resulting solution is free from the heavy
loading of alkali metals winch follows the usual fusion procedures. A full
discussion of decomposition techniques is given in Reference 16.

in, 36. PRFxiPrm ION. The conditions for precipitation are given in
Section XT. 6, Precipitations arc usually carried out in resistance-glass beakers,
and the solution ot the precipitant is added .slowly (for example, by means of a
pipette, burette, or tap funnel) ano with cflicieni stirring of the suitably diluted
solution. The addition must always be made without splashing; tliis is be.st
achieved by allowing the solution of the reagent to flow down the side of the
beaker or precipitating ve.sscl. Only a moderate excess of the reagent is generally
required; a very large excess may lead to increasing solubility (compare Section
II, 8) or contaminatioii of the precipitate. After the precipitate has settled, a few
drops of the precipitant should always be added to determine whether further
precipitation occurs. As a general rule, precipitates arc not filtered off im-
mediately after they have been formed ; most precipitates, with the exception of
those\vhicharedcfinitclycolloidal,suchasiron(ni)hydroxide.rcquiremoreorlcss
digestion (Section XI, 5) to complete the precipitation and make all particles of
filterable size. In some eases digestion is carried out by setting the beaker aside
and leaving the precipitate in contact with the mother liquor at room
temperature for 12-24 hours; in others, where a higher temperature is

106

COMMON APPARATUS AND BASIC TECHNIQUES HI, 37/38

permissible, digestion is usually effected near the boiling point of the solution.
Hot plates, water baths, or even a low flame if no bumping occurs, are employed
for the latter purpose ; in all cases the beaker should be covered with a clock glass
with the convex side turned down. If the solubility of the precipitate is
appreciable, it may be necessary to allow the solution to attain room temperature
before filtration.

in, 37. FILTRATION. This operation is the separation of the precipitate
from the mother liquor, the object being to get the precipitate and the filtering
medium quantitatively free from the solution. The media employed for filtration
are: (1) filter paper; (2) filter mats of purified asbestos (Gooch crucibles) or of
platinum (Munroe crucibles); (3) porous fritted plates of resistance glass, e.g.,
Pyrex (sintered glass filtering crucibles), of silica (Vitreosil filtering crucibles), or
of porcelain (porcelain filtering crucibles) : see Section HI, 28.

The choice of the filtering medium will be controlled by the nature of the
precipitate (filter paper is especially suitable for gelatinous precipitates) and also
by the question of cost. The limitations of the various filtering media are given in
the account which follows.

Ill, 38. FILTER PAPERS. Quantitative filter papers must have a very small
ash content; this is achieved during manufacture by washing with hydrochloric
and hydrofluoric acids. The sizes generally used are circles of 7.0, 9.0, 1 1.0, and
12.5 cm diameter, those of 9.0 and 1 1.0 cm being most widely employed. The ash
ofa 1 1-cm circle should not exceed 0.000 1 g ; if the ash exceeds this value, it should
hededucted from the weight of the ignited residue. Manufacturers give values for
the average ash per. paper: the value may also be determined, if desired, by
Igniting several filter papers in a crucible. Quantitative filter paper is made of
various degrees of porosity. The filter paper used must be of such texture as to
retain the smallest particles of precipitate and yet permit of rapid filtration. Three
textures are generally made, one for very fine precipitates, a second for the average
precipitate which contains medium-sized particles, and a third for gelatinous
precipitates and coarse particles. The speed of filtration is slow for the first, fast
f°rtb medium for the second. ‘Hardened’ filter papers are made by

j I ^'’^matment of quantitative filter papers with acid ; these have an extremely
to ^ -a ^ greater mechanical strength when wet, and are more resistant
Th ^^ey should be used in all quantitative work,

man p™®’^^^^eristics of the Whatman series of hardened ashless filter papers,
consult ^ shown in Table III, 7: for further details

- ^^>7 Whatman’ quantitative filter papers

Fflter

paper

Fast speed.

Medium

Slow speed.

Retains

speed.

Retains

coarse

Retains

fine

particles

medium-sized

particles

washed, hardened

No. 541

0.08

No. 540

0.08

No. 542
0.09

107

ni, 38 OUANTITATIVK INORGANIC ANALYSIS

The size of the filter paper selected for a particular operation is dcicrinincd by
the bulk of the precipitate, and not by the %-olumc of the liquid to be filtered. The
entire precipitate should occupy about a third of the capacity of the filter at the
end of the filtration. Tlic funnel should match the filter paper in size; the folded
paper should extend to within l-.2cin of the top of the funnel, but never closer
than 1 cm.

A funnel with an angle as nearly 60' as po.ssible should be employed ; the stem
should have a length of about 1 5 cm in order to promote rapid filtration. The
filter paper must be carefully fitted into the funnel so that the upper portion beds
tightly against the glass. Some analy.sts recommend that the filter paper should
rest completely against the wall of the funnel ; this is really unncccssaty. since a
filter paper which adheres snugly to the funnel over the upper half only will
permit more rapid filtration. To prepare the filter paper for u.se. the dry paper
is usually folded exactly in half and exactly again in quarters. The folded paper is
then opened so (hat a 60' cone is formed vv-ilh three lliickncsses of paper on the
one side and a single thickness on the other; the paper is then adjusted to fit the
funnel. The paper is placed in the funnel, moistened thoroughly with water,
pressed down lightly to tite sides of the funnel, and then filled with water. If the
paper fits properly, the stem of (lie funnel will remain filled with liquid during the
filtration. Another method of folding tiic filter, which is preferable to that just
described, consists in folding the paper acro.ss a diameter and then once again so
that (he two halves of the first crease do not quite coincide (the two cxtrcmcedgcs
should enclose an angle of 3- 4' for a 60 funnel) ; t he corner of the fold should be
torn ofi'to a depth of about one-third of the radius of the paper. When this filter is
opened and placed in the funnel, it shouUi fit the walls tightly at the upper half: if
it docs not fit properly, the angle of the second fold must be adjusted until it docs,
Here also, the ultimate test of a proper fit is that the stem of the funnel remains
filled with liquid throughout the filtration.

To carry out a filtration, the funnel containing the properly fitted paper is
placed in a funnel stand (or is supported vertically in some other way) and a cle.an
beaker placed so that (he stem of the funnel just touches (beside; this will prevent
splashing. Hie liquid to be filtered is then poured down a glass rod into the filter,
directing the liquid against the side of the filter and not into the apex (Fig. Ill,
27); the lower end of the stirring rod should be vety close
to, but should not quite touch, the filter paper on the side
having three thicknesses of paper. The paper is never filled
completely with the solution ; the level of the liquid should
not rise closer than to within 3-10 mm from the top of the
paper. A precipitate which tends to remain in the bottom
of the beaker should be removed by holding the glass rod
across the beaker, lilting the beaker. ;ind directing a jet
of water from a wash bottle so that the precipitate is
rinsed into the filter funnel. This procedure may also be
adopted to transfer the last traces of the precipitate in the
beaker to the filter. Any precipitate which adheres firmly to
the side of the beaker or to the stirring rod may be
removed with a rubber-tipped rod or ‘policeman' (Section
111,27).

Filtration by .suction is rarely necessary': with gela-
tinous and some finely divided precipitates, the suction will

108

COMMON APPARATUS AND BASIC TECHNIQUES HI, 39/40

draw the particles into the pores of the paper, and the speed of filtration will
actually be reduced rather than increased. If suction is used with filter paper, it is
necessary to support the paper in a perforated cone made of platinum (‘filter
cone’) or in a Whatman filter cone (hardened, No. 51).

in, 39. FILTER PULP. Dittrich first recommended the use of filter paper
pulp as an aid in the filtration and washing of gelatinous or slimy precipitates,
which tend to clog the pores of ordinary filter paper. The macerated paper may be
prepared by vigorously shaking an ordinary quantitative filter paper, torn into
small pieces, with hot distilled water in a stoppered conical flask until it is
disintegrated to a pulp. Ashless grade filter clippings (ash not exceeding 0. 1 per
cent) are marketed (Whatman) for the preparation of pure filter pulp by
dispersing in distilled water.

Filter pulp tablets are marketed ready for use; and are easily disintegrated in
distilled water. Whatman ‘accelerators’ are small discs, each weighing ca. 0.4g and
giving an ash of about 0.00006 g. Whatman ‘ashless tablets’ are larger, weighing
2.4geach, and giving an ash of about 0.0003 g. Either one or two ‘accelerators’ or
a quarter of a ‘tablet’ is used in the average precipitation. The bulk of the filter
pulp should be approximately equal to that of the gelatinous precipitate. The
fflter pulp is added after the precipitate has formed and immediately before
^ration. When the bulk of the precipitate and paper is large, it is usually
advisable to support the filter paper on a Whatman filter cone and drain well on
the pump upon completion of the washing. It is best not to dry the filter and
precipitate completely, as a hard mass may be formed, which is difficult to insert
in he crucible in the subsequent ignition. While still slightly moist it should be
ransferred to the crucible, and the drying completed in the crucible by heating
over a very small non-luminous flame.

ffi)40. GOOQI CRUCIBLES. The characteristics of Gooch crucibles have
hnlrt Section III, 28. In use, the crucible is supported in a special

or, nown as a Gooch funnel, by means of a wide rubber tube (Fig. Ill, 28) ;

the bottom of the crucible should be quite
free from the side of the f unnel and from the
rubber gasket, the latter in order to be sure
that the filtrate does not come into contact
with the rubber. The Gooch funnel passes
through a one-holed rubber bung into a
large filter flask of about 750 cm^ capacity.
The tip of the funnel must project below the
side arm of the filter flask so that any risk
that the liquid may be sucked out of the
filter flask may be avoided. The filter flask
“‘“luarcana 't should be coupled with another flask of

M latter connected to a water filter pump ; if the water in

’’“I be conta°^' ° back’, it will first enter the empty flask and the filtrate will
lo limit the advisable also to have some sort of pressure regulator

niethod is t pressure under which filtration is conducted. A simple

2ltemativelv° ^ ™ the„second filter flask, as in Fig. Ill, 28;

and ® T-piece may I f -'iced between the receiver and the
one arm closed eitheri ' ' lap or by a piece of heavy rubber

109

Ill, 41 QUANTITATIVE INORGANIC ANALYSIS

tubing ('pressure' tubing) carrying a screw clip.

The rubber sleeve for fitting the Gooch crucible into the Gooch funnel may be
replaced with advantage by either;

((/) a .solid rubber ring, shaped to hold the crucible, which fits into a 7.5-cm filter
funnel ;

(/?) a solid rubber washer; the wide part of the washer merely rests on top of the
Gooch funnel, and the inside part conforms to the shape of. and supports the
crucible.

The Gooch cruciblcshould be of suitable si/e. One about 4 cm in height, with a
capacity of 25 cm^ and perforations about 0.5-0.8 mm in diameter, will be found
serviceable for most purposes, The crucible is first placed in the suction-filtering
apparatus, and then half to two-thirds filled with the suspension of asbestos in
water. The whole is allowed to stand for 2-3 minutes in order to allow the larger
particles to settle to the bottom, and then suction is applied gently. When the
water has passed through, the pump is turned on full, and the mat .sucked down
tight. The final uniform pad of asbestos should have a thickness of 2-5 mm: it is
possible to tell appro.ximntcly when the mat has the correct thickness by holding
the crucible up to the light and looking through it. when the outline of the holes
should be barely visible. If the pad is too thin, more asbestos must be added and
the process repeated. The asbestos pad is now iluuoughly washed with distilled
water under the ma.ximum suction of the pump until no fine fibres pa.ss into the
filtrate. It should be mentioned that some analysts prefer to place a perforated
porcelain plate (‘Witt’ plate) upon the asbestos mat to prevent its dislodgcmcnt:
a little more of the suspension is poured in to furnish enough asbestos to barely
cover the plate and to hold it in place. This procedure is unna'cssary if it be
remembered that no liquid may be poured into the crucible unless suction is being
applied. Liquids should be poured gently on to tlic centre of the mat down a
stirring rod ; a jet of water from a wash bottle should never be directed into the
prepared cnicibic. If these precautions are taken there is* little danger that the mat
wall become torn and allow the precipitate to pass through.

The crucible is placed on a small ignition dish or saucer or upon a shallow-
form Vitreosil capsule and dried to constant weight at the same temperature as
that which will be subsequently used in drying tlic precipitate. For temperatures
up to about 250 C a thermostatically controlled electric oven should be used.
For higher tcmperature,s, the crucible may be heated in an electrically heated
mufile furnace. In all eases the crucible is allowed to cool in a desiccator before
weighing. The asbe-stos normally used for Gooch crucibles tends to lose weight
above about 280‘C. hence it is recommended that precipitates which require
heating above about 250 'C should not be collected in Gooch crucibles. Porous
filtering crucibles (sec below) may be employed.

in, 41. aiUCIBLES WITH permanent porous PLATES. Ref-
erence has already been made to these crucibles and to cruciblcs^with a
porous base in Section III, 28. They possess an obvious advantage over Gooch
crucibles in liiat no preparation of a filter mat is necessary, and there is none
of the possible instability associated w'itli filter mats, i.c., the possibility that the
mat niay become dislodged during the filtering process. They are used c.xactly as
described for Gooch crucibles, and a Gooch funnel is used to support the
crucible during the filtration process.

Care must he taken with porous base crucibles, as with sintered glass crucibles,

COMMON APPARATUS AND BASIC TECHNIQUES IH, 42

to avoid attempting the filtration of materials that may clog the filter plate. A new
cracible should be washed with concentrated hydrochloric acid and then with
distilled water. The crucibles are chemically inert and are resistant to all solutions
which do not attack silica; they are attacked by hydrofluoric acid, fluorides, and
strongly alkaline solutions. ,

Crucibles fitted with permanent porous plates are cleaned by shaking out as
much of the solid as possible, and then dissolving out the remainder of the solid
with a suitable solvent.

A hot O.IM solution of the tetrasodium salt of ethylenediamine tetra-acetic
add is an excellent solvent for many of the precipitates (except metallic sulphides
and hexacyanoferrates(III)) encountered in analysis. These include barium
sulphate, calcium oxalate, calcium phosphate, calcium oxide, lead carbonate,
leadiodate, lead oxalate, and amriionium magnesium phosphate; for calcium
and barium precipitates the solution must be alkaline (pH >10). The crucible
may either be completely immersed in the hot reagent or the latter may be drawn
by suction through the crucible.

in, 42. WASHING OF PRECIPITATES. Most precipitates are produced in
the presence of one or more soluble compounds. Since the latter are frequently
not volatile at the temperature at which the precipitate is ultimately dried, it is
necessary to wash the precipitate to remove such material as completely as
popible. The minimum volume of the washing liquid required to remove the
objectionable matter should be used, since no precipitate is absolutely insoluble.
Qualitative tests for the removal of the impurities should be made on small
volumes of the filtered washing solution. Furthermore, it is better to wash with a
number of small portions of the washing liquid, which are well drained between
oach washing, than with one or two large portions, or by adding fresh portions of
a washing liquid whilst solution still remains on the filter (see Section XI, 8).

he ideal washing liquid should comply as far as possible with the following
conditions: j t'

■ it should have no solvent action upon the precipitate, but dissolve foreign
substances easily;

3 ii dispersive action on the precipitate ;

4 'f 1 volatile or insoluble product with the precipitate ;

5 it I volatile at the temperature of drying of the precipitate ;

• s ould contain no substance which is likely to interfere with subsequent
j terminations in the filtrate.

dissoP”^^^^’ water should not be used unless it is certain that it will not
nPDre'^’^ appreciable amounts of the precipitate or peptise it. If the precipitate is
clectrd^ ^ ^°^^We in water, a common ion is usually added, since any
Pbre Wat ^ ^ .soluble in a dilute solution containing one of its ions than it is in

dilute a H, 9) ; as an example the washing of calcium oxalate with

become *^7)° oxalate solution may be cited. If the precipitate tends to
with geH' through the filter paper (this is frequently observed

clectrolvt* flocculent precipitates), a wash solution containing an

c'ectrolvt^ employed (compare Section XI, 3). The nature of the

"'ashing a^ H provided it has no action upon the precipitate during

^^lected J® yolatilised during the final heating. Ammonium salts are usually
'"ashing iro thus ammonium nitrate solution is employed for

"I I) hydroxide. In some cases it is possible to select a solution which

Ill, 43/44 QUANTITATIVE INORGANIC ANALYSIS

will both reduce the solubility of the precipitate and prevent peptisation; for
example, the use of dilute nitric acid witli silver chloride. Some precipitates tend
to oxidise during washing ; in such instances the precipitate cannot be allowed to
run dry, and a special washing solution which rc-converts the oxidised
compounds into the original condition must be employed, c.g., acidulated
hydrogen sulphide water for copper sulphide. Gelatinous precipitates, like
aluminium hydroxide, require more washing than crystalline ones, such as
calcium oxalate. With gelatinous precipitates there is also a danger of channel
formation if the wash liquid is allowal to drain completely; these precipitates
should be washed as far as possible by decantation.

in, 43. TECHNIQUE OF FILTRATION. When the proper filtering
medium (filter paper, Gooch crucible, etc.) has been prepared, as much as
possible of the supernatant liquid is poured off by directing the stream ofliquid
against a glass rod held against the lip of the beaker (compare Fig. Ill, 27. Section
ifl, 38) wthout disturbing the precipitate. The precautions already mentioned
against filling a filter paper too full must be observed. In most ca.scs, particularly
if the precipitate settles rapidly or is gelatinous, washing by decantation may be
employed. Twenty to fifty cm' of a suitable wash liquid is added to the residue in
the beaker, the solid stirred up and allowed to settle. If the solubility of the
precipitate allows, the solution should be heated, since, iniiT alia, the rate of
filtration will thus bo increased. When the supernatant liquid is clear, as much as
possible of the clear liquid is decanted through the filtering medium. This process
is repeated three to five times (or as many time* as is necessary') before the
precipitate is transferred to the filter. The main bulk of the precipitate is first
transferred by mixing with the wash solution and pouring olT the suspension, the
process being repeated until most of the solid has been remos ed from the beaker.
The precipitate adhering to the sides and liic bottom of the beaker is removed as
follows. The beaker is grasped in the left hand. ;ind the stirring rod is held firmly
against the top of the beaker with the index finger and should project 2-3 cm
beyond the lip; the wash bottle is controlled by the right hand. The beaker is
inclined and a stream of water (or wash liquid) is directed against the precipitate
to dislodge it and wash it against the rod into the filter or filtering crucible. After
the above treatment there will generally be small amounts of the precipitate
adhering to the walls of the beaker. These are removed by rubbing w'ith a
‘policeman’; when all the particles have been dislodged, the ‘policeman' is rinsed
with the wash liquid, and the remaining precipitate transferred to the filter or
filtering crucible.

Where the precipitate is washed on the filter, in the last stagc-s the washing
solution is directed along the rim and then gradually towards the apex of the
cone. In all eases, tests for the completeness of washing must be made by
collecting a small sample of the washing solution after it is estimated that mo.st of
the impurities have been removed, and applying an appropriate qualitative test.
Where filtration is carried out under suction, a small test-tube may be attached to
the bottom of the Gooch funnel by means of a wire.

Ill, 44. DRYING AND IGNITION OF FREClPITATFis. After a pre-
cipitate has been filtered and washed, it must be brought to a constant
composition bclorc it can be weighed. The further treatment will depend both

112

COMMON APPARATUS AND BASIC TECHNIQUES III, 44

« the nature of the precipitate and upon that of the filtering medium; this
iLnlent consists in drying or igniting the precipitate. Which of the latter two
terms is employed depends upon the temperature at which the precipitate is
lieated There is, however, no definite temperature below or above which the
precipitate is said to be dried or ignited respectively. The meaning will be
adequately conveyed for our purpose if we designate drying when the tempera-
te is below 250 °C (the maximum temperature which is readily reached in the
Bsual thermostatically controlled, electric drying-oven), and ignition above
250'Cupto,say, 1200 °C. Precipitates that are to be dried should be collected on
ier paper, or in Gooch, sintered glass, or porcelain filtering crucibles.
Precipitates that are to be ignited are collected on filter paper, porcelain filtering
crucibles, or silica filtering crucibles. Ignition is simply effected by placing m a
special ignition dish or in a larger nickel or platinum crucible, and heating with
the appropriate burner; alternatively, these crucibles (and, indeed, any type of
crucible) may be placed in an electrically heated muffie furnace, which is
equipped with a pyrometer and a means for controlling the temperature.

Attention is directed to the information provided by thermogravimetric
analysis (see Chapter XXIII) concerning the range of temperature to which a
predpitate should be heated for a particular composition. In general, thermal
gravimetric curves seem to suggest that in the past precipitates were heated for
too long a period and at too high a temperature. It must, however, be borne in
mind that in some cases the thermal gravimetric curve is influenced by the
experimental conditions of precipitation, and even if a horizontal curve is not
obtained, it is possible that a suitable weighing form may be available over a
certain temperature range. Nevertheless, thermograms do provide valuable data
concerning the range of temperature over which a precipitate has a constant
composition under the conditions that the thermogravimetric analysis was
mode; these, at the very least, provide a guide for the temperature at which a
precipitate should be dried and heated for quantitative work, but due regard
must be paid to the general chemical properties of the weighing form.

Although precipitates which require ignition will usually be collected in
porcelain or silica filtering crucibles, there may be some occasions where filter
Poper has been used, and it is therefore necessary to describe the method to be
0 opted in such cases. The exact technique will depend upon whether the
precipitate may be safely ignited in contact with the filter paper or not. It must be
remembered that some precipitates, such as barium sulphate, may be reduced or
*• ^oged in contact with filter paper or its decomposition products.

•u of the filter paper in the presence of the precipitate. A silica

le is first ignited to constant weight (i.e., to within 0.0002 g) at the sarne
that to which the precipitate is ultimately heated. The well-
funn'h paper and precipitate are carefully detached from t e

bein ^ 1 paper is folded so as to completely enclose the precipitate, care
wp! n j paper. The packet is then placed point down in the

resHn^ ‘'^cible, which is supported on a pipe-clay, or better, a silica triang e
in tb ^ Fig- IIF 29. The crucible is slightly inclined, as shown

trianot partially covered with the lid, which should rest partly on the

Procppu ^ flame is then placed under the crucible lid ; drying thus

'be without undue risk. When the moisture has bee^expelle^

siigiiiiy so as to slowly carbonise the paper. The paper
be allowed to inflame, as this may cause a mechanical expulsion ot hne

113

in, 44 QUANTITATIVE INORGANIC ANALYSIS

particles of liic precipitate owing to the rapid "
escape of the products of combustion; if, by chance,;' '
it does catch fire, the flame should be extinguished by/ ]
momentarily placing the cover on the mouth of the cru-;‘
cihic with the aid of a pair of crucible tongs. When thee
paper has completely carboni.scd and vapours arc no-
longer evolved, the flame is moved to the back (boN ,
tom) of the crucible and the carbon slowly burned ofT.. .
whilst the flame is gradually increased.* After all the .
carbon has been burned away, the crucible is covered .
completely (if desired, the crucible may be placed in a .
vertical position for this purpose) and heated to the
required temperature by means of a Bunsen or Meker
flame. Usually it takes about 20 minutes to char the
paper, and 30-60 minutes to complete the ignition.
Wlren the ignition is ended, the flame is removed .and. ,
after 1-2 minutes, the crucible and lid arc placed in a desiccator containing a
suitable desiccant (Section HI, 26). and allowed to cool for 25-30 minutes. The
crucible and lid are then weighed : The crucible and contents arc then ignited at ;
the same temperature for 10-20 minutes, allowed to cool in a desiccator as
before, and weighed again. The ignition is repeated until constant weight is
attained. Crucibles should always be handled with clean crucible tongs and
preferably with platinum-tipped tongs.

It is important to note tli.at 'heating toconstant weight' has no real significance
unless the periods of heating, cooling of the crnircr/ crucible. ;ind weighing arc
duplicated. There is some doubt as to whether cooling in a desiccator containing
a desiccant is really successful in all eases in preventing some moisture being
absorbed by the crucible and contents; this possible error is minimised by
covering the crucible and weighing the precipitate as soon ii.s it acquires the
laboratory temperature. The empty crucible and lid should, of course, be
subjected to the same treatment.

B. Incineration of the filter paper apart from the precipitate. This method is
employed in all those eases where the ignited substance is reduced by the burning
paper; for example, barium sulphate, lead sulphate, bismuth oxide, copper
oxide, etc. The funnel containing the precipitate is covered by a piece of
qualitative filter paper upon which is written the formula of the precipitate and
the name of the owner; the paper is made secure by crumpling its edges over the
rim of the fimncl so that they will engage the outer conical portion of the funnel.
The funnel is placed in the steam oven, or in a drying oven maintained at
100-105 ”C, for 1-2 hours or until completely dry. A sheet ofgla'/cd paper about
25 cm square (white or black, to contrast with.thc colour of the precipitate) is
placed on the bench away from all draughts. Tlie dried filter is removed from the
funnel, and as much as possible of the precipitate is removed from the paper and
allowed to drop on a clock glass resting upon the glazed paper. This is readily
done by very gently rubbing the sides of the filter paper together, when the bulk of
the precipitate becomes detached and drops upon the clock glass. Any small

If ihc carbon on tbclntisoxidiscdonty slowly, thcco\cr may be hc.'Ucd scparalcly in a Itamc. It is, of
course, held in clean crucible tongs.

114

COMMON APPARATUS AND BASIC TECHNIQUES III, 45/46

n»riicles of the precipitate which may have fallen upon the ^azed paper are
Eu hed into the crucible with a small camel-hair brush The clock glass
Sai the precipitate is then covered with a larger clock glass or with a
teler The filter paper is now carefully folded and placed inside a weighed
Lainorsilicacracible. The crucible is placed on a triangle and
kinerated as detailed above. The crucible is allowed to cool, and the filter ash
rfccted to a suitable chemical treatment in order to convert any reduced or
clanged material into the form finally desired. The cold crucible is then placed
tpon the glazed paper and the main part of the precipitate carefully transferred
from the clock glass to the crucible. A small camel-hair brush will assist in the
transfer. Finally, the precipitate is brought to constant weight by heating to the
necessary temperature as detailed under A.

in, 45. PERFORATED SCREENS FOR CRUCIBLES. It is (often import-
ant to exclude flame gases from the interior of a crucible during an ignition, e.g..
in the ignition of iron(IIl) oxide. For this purpose we may employ a vitreosil
plate, about 10cm square, in which a round opening is cut large enough to admit
Ihecrucible to two-thirds of its depth. The plate is held at an angle of about 30
from the horizontal by means of a clamp; alternatively, but less satisfactorily, it
may be suspended on a tripod. Asbestos board may also be employed, but this
has the disadvantage that fibres may adhere to the crucible: this difficulty is less
liiely to occur with ‘Uralite’.

ID, 46. THE SCHONIGER OXYGEN FLASK METHOD FOR
elemental ANALYSIS. One of the most useful methods available for
micro-analysis is that developed by Schoniger in 1955 (Refs. 17 and 18 ).Itisbase<i
upon the procedure for the combustion of organic materials in an atmosphere of
originally introduced by Hempel (Ref. 1 9) for determining sulphur in coal.
Although the procedure is used to analyse organic substances it is included in
mis book as the elements of the organic materials combusted are, in fact,
fterrained in their inorganic forms using many of the titrimetric or spectro-
Pnotomeric methods described in later sections.

J?!!™!’®.'' of reviews of the oxygen flask method have been published (Refs. 20
) giving considerable details of all aspects of the subject.

D outline the procedure consists of carefully weighing about 5-10 mg o
a shaped piece of paper (Fig. Ill, 30c) which is folded in such a way
(lien j (wick) is free. This is then placed in a platinum basket or earner
ground glass stopper of a 500 cm^ or 1 litre flask. The flask,
fillM ^ absorbing solution (e.g. aquec • 'um hydroxide), is

attach"^!)^ and then sealed with the stopper :• , datinum baske

Ill, 47 QUANTITATIVE INORGANIC ANALYSIS

ia) (f’) 'T;

Fig. HI, 30 Conventional flasks for mlcrodcfcrniiiiations: (a) airleak deslRii, (b)
stopper dcsipn (c) filter paper for wrappine sample
Reproduced by permission from A. M. f». Macdonald (1965). In
AdraTtet'i in Anniyiicitl Chemhiry nnit imlrumentalhn. {Ed. C. N.

Reilley), )'ol 4, p. 75. New )'orK; Inferscience.

combustion, absorbed in liydrogcn peroxide, and the sulphur determined as
.sulphate.

The combustion products of organic halide.s arc usually absorbed in sodium
hydroxide containing some hydrogen peroxide. The resulting solutions may be
analysed by a range of available prevcedurcs. For cliloridcs the method most
commonly used is that of argentimctric poicntiometric titration (Ref. 22) (see
Section XIV, 25). whilst for bromides a merciirimclric titration (Ref. 23) is
comparable with the argentimctric method.

Phosphorus from organophosphorus compounds, which arc combusted to
give mainly orthophosphate, can be absorbed by cither sulphuric acid or nitric
acid and readily determined spcctrophotometrically cither by the molybdenum
blue method or as the pho.sphovanadomolybdatc (Section XVJII, 33).

Procedures have also been devised for the determination of metallic con-
stituents. Thus, mercury is absorbed in nitric acid and titrated with sodium
dicthyldithiocarbamatc, whilst zinc is atv^orbed in hydrochloric acid and
detennined by an EDTA titration (see Section X, 67).

in, 47. References

1. Nation.ii Physic.al Laboratory (1956) Dalamrs. nV(g/;f.r nnd Pnxhf luihnratory
Weighing. London; Her Majesty’s Stationery Oniee.

2. (a) J. T. Stock (1969). The Dcvchpmau of the Chriitical Balance. London; Her
Majcsty’.s Stationery Office.

(b) J. T. Stock (1973). Analytical Chemistry. 45, 947A.

3. (a) T. W. Richards (1900). J. Am. Chem. Sac.. 22, 144.

(b) A. 1. Vogel (1961). Quantitative Jiwrqanic .-inalrsis. 3rd cdn. London; Longmans.

4. T. W. Lashof and L. IL MacCurdy (1954). Analytical Chemistry. 26, 707.

5. M. L. McGIa.shan (1971). Physico-Chemical Qmintitics and Units, 2nd cdn (p. 45).
London; Royal Institute of Chemistry.

6. C. Woodward and 1 1. N. Redman ( 1 973). Iliyh Precision Titrimctry. London ; Society
for Analytical Cliemistry.

(a), p. 14.(b),p.5.(c).p. lL(d).p. 10. .and p. 12,

7. Technical Information Bulletin (1967). Gnnimetric Titrimctry— A Review of the
Literature. Princeton ; Metticr Instrument Corptiration.

116

COMMON APPARATUS AND BASIC TECHNIQUES III, 48

! (alBS3978 1966: Water for Laboratory Use. London; British Standards Institution.
'([,) 01193-70 {1970). Standard Specification for Reagent Water. Easton, Md.,

Washington, DC; American Chemical
Tiw'Manipulation of Air-Sensitive Materials. New York,

McGraw-Hill Book Co. „ . -n n

IJ,RG.BatesandE.Wichers(1957).ARcj.Nflr/.BMr.S/flHrf.,59,9.

II. EB.Sanddl and T. S. West (1969). ‘Recommended Nomenclature for Titnmetnc

Mpi.’Pureand Applied Chemistry, IS, 429. u • a i •

II B.Beraas (1968). ‘A New Method for Decomposition and Comprehensive Analysis
ofSilicates.’/lna/. C/ie»i.,40, 1682.

LlJ.C.MeakinandM.C. Pratt(1972). Manual of Laboratory Filtration. Maidstone; b.

W. Baton Ltd.

H, C.0.1ngameUs(1964). ‘Rapid Chemical Analysis of Silicate Rocks.’ Talanta, 11, 665.
15. H. Bennett and G. J. Oliver (1971). ‘Loss of Cobalt and Iron from Lithium Borate
Fusions in Graphite Crucibles.’ /tna/yjt, 96, 427.

Ifi. J. Dolezal, P. Povondra and Z. Sulcek (1968). Decomposition Technicpies in Inorganic
Mysis. London; Iliffe Books Ltd.

II. W.Sch6niger(1955). Mikrochim. Acta, 123.

18. W.Sch6niger(1956). Mikrochim. Acta, 869.

19. W.Hempel(1892). Z. Angew. Client., 13, 393.

20. A. M.G. Macdonald (1965), in C. N. Reilley (ed.). Advances in Analytical Chemistry
enllmtrwnentation. Vol. 4, p. 75, New York; Intersciencc.

21. A. M.G. Macdonald (1961). Analyst, 86, 3.

22. Analytical Methods Committee (1963). Analyst, 88, 415.

23. R.C.Denney and P. A. Smith (1974). Analyst, 99, 166.

2A Asbestos Regulations (1969), Department of Employment Health Booklets and
Technical Data. H.M.S.O., London.

11)48. Selected bibliography

'hay of the books detailed in Section 1, 12 contain descriptions of simple apparatus and
® techniques and are relevant to the present Chapter. Attention is particularly directed

bl tt” Elving. Treatise on Analytical Chemistry (Vols. 7, 9, 1 0).

cl Parry Jones. Chemical Analysis: the Working Tools.

I ^(’f^Pi'chensive Analytical Chemistry (Voi.lA).

• w. M MacNevin (1951). The Analytical Balance. Its Care and Use. Handbook
Publishers, Sandusky, Ohio.

■ Physical Laboratory (1956). Balances, Weights and Precise Laboratory

londm Applied Science, No. 7.) Her Majesty’s Stationery Office,

■ H. Bigg ( 1959 ) ‘Weight-in-air Basis of Adjustment of Precision Weights. Journal
4 R^v 36, 359.

BS S Institution, London. .

ko/«me(f Metric Units of Volume and Standard Temperature of

BS604

BS676

BS700

BS846

1q« Measuring Cylinders.

1 QM Graduated Necks.

Q«' Pipettes and One-Mark Cylindrical Pipettes.

iy 62 . Rtirotfnn n..1L n ...

BS846- ripeues am

BS mi , n ■ ^^’’^ftes and Bulb Bure
“^‘ 132 : 1966 .

BS 1583
BS 1792

., ...ico u,iu Duio jiurettes.
*966. Automatic Pipettes.

iQ^fi' Bulb Pipettes.

60. One-Mark Graduated Flasks.

117

Ill, 48 OUANTlTATlVi- INORGANIC ANALYSIS

BS 1797: 1968. Tables for Lhe imhf Calibration of VoUmwtric Glassware.

BS205S; 1961. iVeifthinR Pipettes.

BS 1752; 1963. Laboratory .sintered or fritted filters.

BS 2648: 1955. Performance rcqtdrcntcnts fin elertrieally-heated laboratory drying
osen.s.

BS 3423 ; 1962. Rccommcndationsfor the design of glass vacuum de.siccators.

BS397S: 1966. IVater for laboratory ti.se.

BS 3996: 1966. Colour coding for one-rnark and graduated pipettes.

BS4244; 1967. Porcelain and silica crucibles.

BS 2445 : 1 968. Recommendations for solutions used in chemical analysis. Terminology,
presentation and concentration.

5. R. F. HtRch. ’Modern bibomtory B;tl:inccs. Part I.‘ J. Client. Ed.. 44, A1023 1967
■Part II.V. C/«w. £r/.. 45. A7. (1968).

6. 'AnalaR' Standards for UilKiraiory Chemicals (1967). 6th edn. London; AnalaR
Standard.^ Ltd.

II8

PART B ERRORS AND SAIVIPLING

CHAPTER IV ERRORS AND STATISTICS

IV, 1. LIMITATIONS OF ANALYTICAL METHODS. The function of
the analyst is to obtain a result' as near to the true value as possible by the correct
application of the analytical procedure employed. The level of confidence that
the analyst may enjoy in his results will be very small unless he has knowledge of
the accuracy and precision of the method used as well as being aware of the
sources of error which may be introduced. Quantitative analysis is not simply a
case of taking a sample, carrying out a single determination and then claiming
that the value obtained is irrefutable. It also requires a sound knowledge of the
c emistry involved, the possibilities of interferences from other ions, elements
and compounds as well as the statistical distribution of values. The purpose of

m j explain some of the terms employed and to outline the statistical
methods that may be applied.

rnLn ^ . The accuracy of a determination may be defined as the

mptlinH oetween it and the true or most probable value. For analytical

■ ahiniiif * possible ways of determining the accuracy ; the so-called

absolve method and the comparative method.

synthetic sample containing known amounts of the
obtaineH K * used. Known amounts of a constituent can be

comnnciti ^ pure elements or compounds of known stoichiometric

merciallv substances, primary standards, may be available com-

Puriiicfltinn^K Prepared by the analyst and subjected to rigorous .

The test of th^ ^^^rystallisation, etc. The substances must be of known purity,
taking varv' ^ ^^‘^uracy of the method under consideration is carried out by
instructiom of the constituent and proceeding according to specified

heterminatp ^ amount of the constituent must be varied, because the
difference Procedure may be a function of the amount used. The

•he constitup"!^^” mean of an adequate number of results and the amount of
'Measure of ^^fually present, usually expressed as parts per thousand, is a
Theconstit^ accuracy of the method in the absence of foreign substances,
of other subst^^^^ question will usually have to be determined in the presence
opon the deter therefore be necessary to know the effect of these

of olements require testing the influence of a large number

•osts may be li^ . ^^rying amounts— a major undertaking. The scope of such
specified range "f o:onsidering the determination of the component in a
0 concentration in a material whose composition is more or less

121

rV, 3 QUANTITATIVE INORGANIC ANALYSIS

fixed both with respect to the elements which may be present and their relative
amounts. It is desirable, however, to study the effect of ns many foreign elements
as feasible. In practice, it is frequently found that separations will be required
before a determination can be made in the presence of varying elements; the
accuracy of the method is likely to be largely controlled by the separations
involved.

Comparative method. Sometimes, as in the analysis of a mineral, it may be
impossible to prepare solid synthetic samples of the desired composition. It is
then ncccssaiy to resort to standard samples of the material in question (mineral,
ore, alloy, etc.) in which the content of the constituent sought has been
detennined by one or more supposedly ‘accurate’ methods of analysis. Tliis
comparative method, involving secondary .standards, is obviously not altogether
salisfactor)’ from the theoretical standpoint, but is nevertheless verj- useful in
applied analysis. Standard samples arc issued by the US National Bureau of
Standards, Washington, and by the Bureau of An.aly.scd Samples,
Middlesbrough.

If several fundamentally different methods of analysis for a given constituent
arc available, c.g.. gravimetric, titnmetric. spcctrophotometric. or .spcctro-
graphic, the agreement between at least two methods of essentially dilferent
character can usually be accepted as indicating the absence of an appreciable
determinate error in cither (a determinate error is one which can be evaluated
experimentally or theoretically).

n’, 3. PRECISION. Precision m:iy be defined :is the concordance of a scries
of measurements of the same quantity. The mean deviation or the relative mean
deviation is a measure of precision. In quantitative analysis the precision of
measurements rarely exceeds 1 to 2 parts per thousand.

Accuracy expresses the correctness of ;i measurement, and precision the
reproducibility of a measurement. Precision always accompanies accuracy, but a
high degree of precision docs not imply accuracy. This may be illustrated by an
example.

Example. A substance was known to contain 49.06 + 0.02 per cent of a
given constituent A. The results obtained by two observers using the same
substance and the same general technique were :

Obscrvcr(I). 49.01 ;49. 21 ;49.0S. Mean 49. 10 percent.

Relative mean error = (49.10-49.06)/49.06 ~ O.OS percent.

Relative mean dcM'ation* = [(0.09 + 0.1 1 +0.02)/3) x 100/49.10
= 0.15 percent.

Observer (2). 49.40; 49.44 ; 49.42. Mean = 49.42 per cent.

Relative mean error = (49.42 - 49.06)/49.06 = 0.73 per cent.

Relative mean deviation* = [(0.02 + 0.02 + 0.00)/3] x 100/49.42
= 0.03 percent.

The analyses of observer (1) were therefore accurate and precise; those of
observer (2) were unusually prcci.se. but less accurate than those of observer (I).
Some small source of constant error appears to be present in the results of (2).

* See Section IV, 7.

122

ERRORS AND STATISTICS IV, 4

IV, 4. CLASSIFICATION OF ERRORS. The errors which affect an
experimental result may be conveniently divided into those of the determinate
and the indeterminate kind.

Determinate or constant errors. These are errors which can be avoided, or
whose magnitude can be determined. The most important of these are ;

1. Operatioriaiand personal errors. These are due to factors for which the
individual analyst is responsible, and are not connected with the method or
procedure; they form part of the ‘personal equation’ of an observer. The errors
are mostly physical in nature and occur when sound analytical technique is not
followed. Examples are: mechanical loss of materials in various steps of an
analysis, underwashing or overwashing of precipitates, ignition of precipitates at
incorrect temperatures, insufficient cooling of crucibles before weighing, allow-
ing hygroscopic materials to absorb moisture before or during weighing, and use
of reagents containing harmful impurities.

Personal errors may arise from the constitutional inability of an individual to
make certain observations accurately. Thus some persons are unable to judge
colour changes sharply in visual titrations, which may result in a slight
overstepping of the end-point.

2. Instrumental and reagent errors. These arise from the faulty
construction of balances, the use of uncalibrated or improperly calibrated
weights, graduated glassware, and other instruments ; the attack of reagents upon
glassware, porcelain, etc., resulting in the introduction of foreign materials;
volatilisation of platinum at very high temperatures; and the use of reagents
containing impurities.

3. Errors of method. These originate from incorrect sampling and from
incompleteness of a reaction. In gravimetric analysis errors may arise owing to
appreciable solubility of precipitates, co-precipitation, and post-precipitation,
decomposition, or volatilisation of weighing forms on ignition, and precipitation
of substances other than the intended ones. In titrimetric analysis errors may
occur owing to failure of reactions to proceed to completion, occurrence of
in uced and side reactions, reaction of substances other than the constituent

eing determined, and a difference between the observed end point and the
stoichiometric end point of a reaction.

4. Additive and proportional errors. The absolute value of an additive
orror is independent of the amount of the constituent present in the de-

rmmation. Examples of additive errors are loss in weight of a crucible in which
snd errors in weights. The presence of this error is
Th ^ r, samples of different weights,
consft^ value of a proportional error depends upon the amount of the
substa ^ proportional error may arise from an impurity in a standard

solution'^n^™^^ leads to an incorrect value for the normality of a standard
constifi f u ^ proportional errors may not vary linearly with the amount of the
urespiii ^r\’ exhibit an increase with the amount of constituent

aluminiu f^^^ple is the ignition of aluminium oxide: at 1200‘^C the
''arious vT' anhydrous and virtually non-hygroscopic; ignition of
type of error appreciably lower temperature will show a proportional

slighuf- accidental errors. These errors manifest themselves by the
observer whvTtif successive measurements made by the same

in the greatest care under as nearly identical conditions as possible.

123

IV, 5 QUANTITATIvn INORGANIC ANALYSIS

They arc due to causes over which the analyst has no control, and which, in
general, are so intangible that they arc incapable of analysis. If a sufficiently large
number of observations is taken, it can be siiown that these errors lie on a cursT of
the form shown in Fig. IV, ! (Section IV, 8). An inspection of this error curs'c
shows; (a) small errors occur more frequently than large ones; (/;) large errors
occur relatively infrequently; and (c) positive and negative errors of the same
numerical magnitude arc equally likely to occur.

IV, 5. MINIMISATION OF ERRORS. Determinate errors can often be
materially reduced by one of the following methods;

1. Calibration of apparatu.s and application of corrcction.s. All instruments
(weights, flasks, burettes, pipettes, etc.) should be calibrated, and the appropriate
corrections applied to the original measurements. In some cases where an error
cannot be eliminated, it is possible to apply a correction for the cflcct that it
produces; thus an impurity in a weighed precipitate may be determined and its
weight deducted.

2. Running a blank determination. This consists in carry'inc out a separate
determination, the sample being omitted, under exactly the .same experimental
conditions as arc employed in the actual analysis of the sample. The object is to
find out the cflcct of the impurities introduced through the reagents and vessels,
or to determine the excess of standard solution necessary to establish the end-
point under the conditions met with in the titration of the unknown sample. A
large blank correction is undesirable, because the c.xact value then becomes
uncertain and the precision of the analysis is reduced.

3. Running a control determination. This consists in carrying out a
determination under as nearly as possible identical experimental conditions upon
a quantity of a standard substance which contains the same weight of the
constituent as is contained in tlic unknown sample. The weight of the constituent
in the unknown can then be calculated from the relation :

Result found for standard _ Wciglit of constituent in standard

Rcsuli found for unknown ~

where .v is the weight of the constituent in the unknown.

In this connection it must be pointed out that standard samples w'hich have
been analysed by a number of skilled analysts arc commercially available. These
include certain primary standards (sodium oxalate, potassium hy-
drogcnphthalatc, arscnic(lll) oxide, and benzoic acid) and ores, ceramic
materials, irons, steels, steel-making alloys, and non-ferrous alloys. All of these
arc obtainable from the US Bureau of Standards, Department of Commerce.
Washington, DC, Many of these arc also available as the ‘British Chemical
Standards' and arc supplied by the Bureau of Analysed Samples, Ltd, Newham
Hall, Middlesbrough, England.

4. Use of independent methods of annly.sis. In some instances the accuracy
of a result may be established by carrying out the analysis in an entirely different
manner. Thus iron may first be determined gravimctrically by precipitation as
iron(lll) hydroxide after removing the interfering elements, followed by ignition
of the precipitate to iron(Ill) oxide. It may then be determined litrimctrically by
reduction to the iron(Il) state, and titration w'ith a standard solution of an
oxidising agent, such ns potassium dicliromatc orcerium(IV) sulphate. Another

124

ERRORS AND STATISTICS IV, 6

example that may be mentioned is the detennination of the strength of a
hydrochloric acid solution both by titration with a standard solution of a strong
base and by precipitation and weighing as silver chloride. If the results obtained
by the two radically different methods are concordant, it is highly probable that
the values are correct within small limits of error.

5. Running of parallel determinations. These serve as a check on the result
of a single determination and indicate only the precision of the analysis. The
values obtained for constituents which are present in not too small an amount
should not vary among themselves by more than three parts per thousand. If
larger variations are shown, the determinations must be repeated rmtil satisfac-
tory concordance is obtained. Duplicate, and at most triplicate, determinations
should suffice. It must be emphasised that good agreement between duplicate and
triplicate determinations does not justify the conclusion that the result is correct;
aconstant error may be present. The agreement merely shows that the accidental
errors, or variations of the determinate errors, are the same, or nearly the same,
in the parallel determinations.

6. Standard addition. A known amount of the constituent being de-
termined is added to the sample, which is then analysed for the total amount of
constituent present. The difference between the analytical results for samples
mth and without the added constituent gives the recovery of the amount of
added constituent. If the recovery is satisfactory our confidence in the accuracy
ofthe procedure is enhanced. The method is usually applied to physico-chemical
procedures such as polarography and spectrophotometry.

7. Internal standards. This procedure is of particular value in spectro-
scopic and chromatographic determinations. It involves adding a fixed amount of
a reference material (the internal standard) to a series of known concentrations of
the material to be measured. The ratios of the physical value (absorption or peak
size) ofthe internal standard and the series of known concentrations is plotted
against the concentration values. This should give a straight line. Any unknown
concentration can then be determined by adding the same quantity of internal
standard and finding where the ratio obtained falls on the concentration scale.

8. AmpMcation methods. In determinations in which a very small amount
ol material is to be measured this may be beyond the limits of the apparatus
available. In these circumstances if the small amount of material can be reacted in
sue a way that every molecule produces two or more molecules of some other
Measurable material the amplification of the quantity may then be within the
“Pe of the apparatus or method available.

■ • dilution. A known amount of the element being determined,
iso? 0 radioactive isotope, is mixed with the sample and the element is
dete pore form (usually as a compound), which is weighed or otherwise
with f redioactivity ofthe isolated element is measured and compared
t>ecalculated^^ odded element : the weight ofthe element in the sample can then

^nL ^^^^®^CANT FIGURES AND COMPUTATIONS. The term digit
digit wh?a°'^^ ‘0” numerals, including the zero. A significant figure is a

digit zero ■ the amount of the quantity in the place in which it stands. The

in the an ^ ?*Sreffoant figure except when it is the first figure in a number. Thus
0-0025 1-2680 g and 1.0062 g the zero is significant, but in the quantity

8 e zeros are not significant figures; they serve only to locate the

125

IV, 6 QUANTITATIVE INORGANIC ANALYSIS

decimal point and can be omitted by proper choice of units, c.g.. 2.5 g. The first
two numbers contain five significant figures, but 0.0025 contains only two
significant figures.

Observed quantities should be noted witJi one uncertain fixture retained. Thus in
most analyses weights are determined to the nearest tenth of a milligram, c.g,,
2. 1 546 g. This means that the weight is Jess than 2. 1 547 g and more than 2. 1 545 g.
A weight of 2.150g would signify that it has been determined to the nearest
milligram, and that the weight is nearer to 2. 1 50 g than it is to cither 2.151 g or
2.149g. The digits of a number which arc needed to express the precision of the
measurement from which the number was derived arc known as significant
figures.

There arc a number of rules for computations with which the student should be
familiar.

1. Retain as many significant figures in a result or in any data as will give only
one uncertain figure. Thus a volume which is known to be between 20.5 cm^ and
20.7 cm^ should be written as 20.6cm^. but not as 20.60cm^. since the latter
would indicate that the value lies between 20.59 cm^ and 20,61 cm*. Also, if a
weight, to the nearest 0.1 mg, is 5.2600 g. it should not be written as 5.260 g or

5.26 g. since in the latter case an accuracy of a centigram is indicated and in the
former a milligram.

2. In rounding oIT quantities to the correct number of significant figures, add
one to the last figure retained if the following figure (which has been rejected) is 5
or over. Thus the average orO.2628, 0,2623, and 0.2626 is 0.2626 (0.26257).

3. In addition or subtraction, there should be in etich number only as many
significant figures as there arc in the least accurately known number. Thus the
addition

168.11+7.045 + 0.6832
should be written

168.11+7.05 + 0.68 = 175.84

The sum or diflcrencc of two or more quantities cannot be more precise than the
quantity having the largest uncertainty.

4. In multiplication or division, retain in each factor one more significant
figure than is contained in the factor having the largest uncertainty. The
percentage precision of a product or quotient cannot be greater than the
percentage precision of the least precise factor entering into the calculation. Thus
the multiplication

1.26 X 1 .236 X 0.6834 x 24.8652

should be carried out u.sing the values

1. 26 X 1.236x0.683x24.87

and the result expressed to three significant figures.

Where a large number of multiplications and divisions are to be made, the use
of logarithms is recommended. Four-figure logarithm tables are sufficiently
precise if interpolation is used.

A 25 cm slide rule is accurate to about 0.25 per cent, and is useful in checking
calculations. The Otis King calculator has an accuracy of about four times that of
the 25 cm slide rule, and the rotary' scales are 170 cm long; it is of convenient size

126

ERRORS AND STATISTICS IV, 7

for the pocket, and is very useful in the analytical laboratory.

With the advent of many reasonably priced electronic pocket calculators
statistical calculations are now easy to carry out and the saving in time achieved
very quickly covers the initial financial outlay. Apart from normal arithmetic
functions a suitablecalculator for statistical work should give squares and square
roots, possess a floating decimal point and at least a six digit display..

For processing large amounts of data, and retrieval or comparison with
previous results, many hours of work are saved by use of computers. Although
computer programming is outside the scope of this book it should be pointed out
that standard programs now exist in Algol, cobol, fortran iv, etc., for
calculating statistical functions and carrying out the more involved mathemati-
cal determinations such as that for binary mixtures by ultraviolet/visible
spectroscopy (Section XVHI, 38).

IV, 7. MEAN (AVERAGE) DEVIATION. STANDARD DEVIATION.
When a quantity is measured with the greatest exactness that the instrument,
method, and observer are capable of, it is found that the results of successive
determinations differ among themselves to a greater or lesser extent. The
average value is accepted as the most probable. This may not always be the
true value. In some cases the difference may be small, in others it may be large;
the reliability of the result depends upon the magnitude of this difference. It is
therefore of interest to enquire briefly into the factors which affect and control
the trustworthiness of chemical analysis.

The absolute error of a determination is the difference between the observed or

measured value and the true or most probable value of the quantity measured.
The absolute error is a measure of the accuracy of the measurement.

The relative error is the absolute error divided by the true or most probable
value; it is usually expressed in terms of percentage or in parts per thousand. The
true or absolute value of a quantity cannot be established experimentally, so that
the observed result must be compared with the most probable value. With pure
substances the quantity will ultimately depend upon the atomic weights of the
constituent elements. Determinations of the atomic weights have been made with
1 e utmost care, and the accuracy obtained usually far exceeds that attained in
or ma^ quantitative analysis; the analyst must accordingly accept their
re lability. With natural or industrial products, we must accept provisionally the
results obtained by analysts of repute using carefully tested methods. If several
analysts determine the same constituent in the same sample by different methods,
e most probable value, which is usually the average, can be deduced from their
esults. In both cases the establishment of the most probable value involves the
PPication of statistical methods and the concept of precision,
dpvi between a series of results is measured by computing their mean

This is evaluated by determining the arithmetical mean of the results,
deviation of each individual measurement from the mean,
finally dividing the sum of the deviations, regardless of sign, by the number
The relative mean deviation is the mean deviation divided by
Ati P he expressed in terms of percentage or in parts per thousand.

example will make this clear.

thesS^^f^ chemistry one of the most common statistical terms employed is
mean ^ population of observations. This is also called the root

quare deviation as it is the square root of the mean of the sum of the

127

IV, 7 QUANTITATIVE INORGANIC ANALYSIS

squares of the difTercnces between the values and (he mean of those values (this is
expressed mathematically below) and is of particular value in connection with
the norma! distribution. Section IV, 8.

Exampk. The percentages of a constituent A in a compound AB were
found to be 4S.32. 48.36, 48.23, 48.1 1, and 48.38 per cent. Calculate the mean
deviation and the relative mean deviation.

Bcsiflis

48.32

48.36

48.23

48.11

48.38

5)241.40

Deviations

0.04

0.08

0.05

0.17

0.10

5)0^4

Mean = 48.28 Mean (kvieition ~ 0.09

Relative mean (kviatinn = 0.09 x 100/48.28 ~ 0.J9 per eeni
= 1.9 parts per thoasatui

If we consider a scries of /; obsers’ntions arranged in ascending order of
magnitude:

.X'l, .Yi, .Yj -Yn. |, .Y„.

the arillinielic mean (often called .simply the mc.an) is given by:

X = •'<ld-X2-b.Yj...-f...-I-.Y„.,-I-.Y„
n

The spread of the values is measured most ctTiciently by the standard deviation .v
defined by:

„ _ ,/(Xi-.y)- + (.v,-.y)-+...{.y„-.y)*

In this equation the denominator is {« - 1 1 rather than n when the number of
values is small.

The equation may also be written as;

s tv-.y f-

n-\ '

The square of the standard deviation is called the variance. A more accurate
measure of the precision, known as the cocOicicnt of variation (C.V.), is given by:

C y _ sx 100

Exampk. Analyses of a sample of iron ore gave the following percentage
values for the iron content : 7.08, 7.2 1 , 7. 1 2, 7.09, 7 . 1 6. 7 . 1 4, 7.07. 7. 1 4. 7. 1 8. 7. 1 1 .
Calculate the mean, standard deviation and cocfTicicnt of variation for the
values.

128

ERRORS AND STATISTICS IV, 8

'results (x)

x-x

(x-X)^

■ 7.08

-0.05 .

0.0025

7.21

. 0.08

0.0064

7.12

-0.01

0.0001

7.09 ..

, ■ -0.04

0.0016

7.16

0.03

0.0009

7.14

0.01

0.0001

7.07

-0.06

0.0036

7.14

0.01

0.0001

7.18

0.05

0.0025

7.11

-0.02

0.0004

total 71.30

E = 0.0182

mean (.x) 7.13

_ /0.0182
V 9 •

= ±0.045
0.045x 100
7.13

= 0.63

IV, 8. NORMAL (GAUSSIAN) DISTRIBUTION. Continuous data, of the
type resulting from a number of analyses of an individual chemical sample, fall
within a range of values that satisfy the Normal (or Gaussian) distribution. This
IS a bell-shaped curve that is symmetrical about the mean as shown in Fig. IV, 1 .

The curve satisfies the
1

y — ^ e

equation :

129

IV, 9 QUANTITATIVE INORGANIC ANALYSIS

With this type of distribution about 68 per cent of all values will fall within one
standard deviation on either side of the mean, 95 per cent will fall within two
standard deviations, 99.7 per cent within three standard deviations, and 99.994
within four standard deviations.

It is important to know that the Greek letters a and /i refer to the standard
deviation and mean respectively of a total population ; whilst the Roman letters s
and .v are used for samples of populations.

IV, 9. COMP.ARISONOFRESUI/fS. Statistical figures obtained from a set
of results arc of limited value by themselves. It is only by comparing them with
the true value or with othersets of data that it is possible to determine whether the
analytical procedure lias been accurate or precise or if it is superior to another
method. There arc three common methods for testing results; (o) Student’s
/-test, {b) the Variance ratio test (/-'-test) and (c) the Chi square distribution.

These methods of test require a knowledge of what is known as the number of
degrees of freedom. In statistical terms this is tlie number of independent values
necessary to determine the statistical quantity. Thus a sample of n values has ti
degrees of freedom, whilst thcsumSit.v — .V)‘ is considered to have/; — 1 degrees of
freedom, as for any defined val tie of.v only n— 1 values can be freely assigned, the
nth being automatically defined from the other values.

(a) Student’s /-test. This is a test (Ref. I ) used for smtill samples ; its purpose is
to compare the mean from a sample with some standard value and to c.xpress
some level of confidence in the significance of the comparison. It is also used to
test the difference between the means of two sets of data .Vi and .\%.

The value of / is obtained from the equation :

/ . ,i,

where ft is the true value.

It is then related to a set of /-tables (Appendix XV) in which the probability (R) of
the /-value falling within certain limits is expressed, cither as a percentage or as a
function of unity, relative to the number of degrees of freedom.

Example. If X the mean of 12 determinations - S.37, and /; the true
value = 7.91, say whclhcror not this result issignificant ifthc standard deviation
is 0.17.

From equation (i)

(8.37-7.91)^/12

^ - 0.17

= 9.4

From / tables for eleven degrees of freedom (one less than those used in the
calculation)

for/> = 0.10(10%) 0.05(5%) 0.01(1%;)

/=I.S0 2.20 3.11

and as the calculated value for / is 9.4 the result is highly significant. The / table
tells us that the probability of obtaining the difference of 0.46 between the
experimental and true result is less than 1 in 100. This implies that some
particular bias exists in the laboratory procedure.

130

ERRORS AND STATISTICS IV, 9

Had the calculated value for / been less than T.SO then there would have been
no significance in the results and no apparent bias in the laboratory procedure, as
the tables would have indicated a probability of greater than 1 in 10 of obtaining
that value. It should be pointed out that these values refer to what is known as a
double-sided, or two-tailed, distribution because it concerns probabilities of
values both less and greater than the mean. In some calculations an analyst may
only be interested in one of these two cases, and under these conditions the /-test
becomes single-tailed so that the probability from the tables is halved.

{b) F-test. This is used to compare the precision of two sets of data (Ref. 2) ; for
example, the results of two different analytical methods or the results from two
different laboratories. It is calculated from the equation :

N.B. the larger value of s is always used as the numerator so that the value of F
is always greater than unity. The value obtained for F is then checked for its
significance against values in the F table calculated from a Gaussian distribution
(Appendix XVI) corresponding to the number of degrees of freedom for the two
sets of data.

Example. The standard deviation from one set of 11 determinations
Sa = 0.210, and the standard deviation from another 13 determinations was
Sb = 0.641. Is there any significant difference between the precision of these two
sets of results?

From equation (ii)

r_. (0.641)^
( 0 . 210 )"^
F=9.4

0.411

0.044

Using the F tables, we look for the values corresponding to 12 degrees of
freedom for and 10 degrees of freedom for These give us three values :

forF = 0.10 0.05 0.01
F=2.19 2.75 4.30

The first value (2.19) corresponds to 10 per cent probabDity, the second value
(2.75) to 5 per cent probability and the third value (4.30) to 1 per cent probability.

Under these conditions there is less than 1 chance in 100 that these precisions
are similar. To put it another way, the difference between the two sets of data is
highly significant.

Had the value of F turned out to be less than 2.19 then it would have been
possible to say that there was no significant difference between the precisions.

(c) Clu square test (x^). This is used to determine whether or not a set of data
h ^ from a theoretical or defined distribution (Ref. 3)^ that is.

Whether the observed frequencies of an occurrence correspond to the predicted
requencies. Chi square is calculated from the equation :

where O is the observed frequency and E the expected frequency.

..(iii)

131

IV, 10 QUANTITATIVE INORGANIC ANALYSIS

Thus if a coin is tossed 100 times and the tails come up 25 times we would wish
to ascertain if there is any real indication of bias.

Normally we would expect an equal chance of obtaining heads or tails, in this
ease 50 heads and 50 tails.

From equation (iii)

2 (25-50)^ , (75-50)^

^ “ 50 50

2x625
50

= 25

For one degree of freedom the tables give the following values (Appendi.x
XVlIl):

1 % level 6.63
0.1 % level 10.83

The value or25 obtained in the above calculation is well beyond I0.S3 and we
can say that there is a significant bias in the spinning of the coin.

IV, 10. THE NUMBER OF I’ARALLEL DETERMINATIONS. To avoid
unnecessary time and expenditure an analyst needs some guide to the number of
repetitive determinations he needs tocarrj’ out to obtain a suitably reliable result.
He will be aware that the greater the number he carries out the greater the
reliability, but at the same time will know that after a cenain number of
determinations any improvement in precision and accuracy is very small.

Although rather involved statistical methods exist for establishing the number
of parallel determinations, a reasonably good assessment can be made by
establishing the variation of the value for the absolute error A obtained for an
increasing number of determinations.

A =

The value for t is taken from the 95 per cent confidence limit column of the /
tables for n - 1 degrees of freedom.

The values for A arc used to calculate the reliability inlenail L from the
equation;

, _ lOOA

where r is the approximate percentage level of the unknown being determined.
The number of replicate analyses is assessed from the magnitude of the change in
L with the number of determinations.

Example. Ascertain the number of replicate analyses desirable {a) for the
determination of approximately 2 per cent Cl" in a material if the standard
deviation lor determinations is 0.05 1, {b) for approximately 20 percent Cl~ if the
standard deviation of determinations is 0.093.

ERRORS AND STATISTICS IV, 11/12/13

(a) For 2 per cent Cl :

Number of
determinations

A = f .

L- 100^%

DiHerence

12.7x0.051 X 0.71 =0.4599

22.99

4.3x0.051 X 0.58 = 0.1272

6.36 .

16.63

3.2x0.051 x0.50 = 0.0816

4.08

2.28

2.8x0.051 xO.45 = 0.0642

3.21

0.87

2.6x0.051 x0.41 = 0.0544

2.72

0.49

(b) For 20 per cent Cl ;

Number of
determinations

A = l£

Difference

12.7x0.093x0.71 = 0.838

4.19

4.3x0.093x0.58 = 0.232

1. 16

3.03

3.2x0.093x0.50 = 0.148

0.74

0.42

2.8x0.093x0.45 = 0.117

0.59

0.15

2.6x0.093x0.41 = 0.099

0.49

0.10

In (a) the reliability interval is greatly improved by carrying out a third
analysis. This is less the case with (b) as the reliability interval is already narrow.
In this second case no substantial improvement is gained by carrying out more
than two analyses.

This subject is dealt with in more detail by Eckschlager (Ref. 4), and Shewell
(Ref. 5) has discussed other factors which influence the value of parallel
determinations.

IV, 11. THE VALUE OF STATISTICS. Correctly used, statistics are an
essential tool to the analyst. They can, in particular, prevent him from making
hasty judgements on the basis of limited information. It has only been possible at
this stage to give a brief resume of the main statistical techniques that may be
applied. The reader is advised to make himself fully conversant with these
methods by obtaining one of the many excellent statistics texts now available.

rv, 12. References

1. Student’ (1908). (W. G. Gosset), Biometrika. 6, 1.

• J. Mandel (1964). The Statistical Analysis of Experimental Data, New York;
Interscience.

3. C J. Brookes, I. G. Betteley and S. M. Loxston (1966). Mathematics and Statistics for
Chemists, New York ; John Wiley, p. 304.

■ K. Eckschlager (1969). Errors, Measurements and Results in Chemical Analysis
London; Van Nostrand Reinhold.

■ C. T. Shewell (1959). /lnfl/yn'ca/C/ie/«rsto’, 31, No. 5, 21A.

133

IV, 13 QUANTITATIVE INORGANIC ANALYSIS

IV, 13. Selected bibliography

1. G. T. Wernimont (1949). ‘Siatislics Applied to Analysis’, Analyrical Chemistry, 21,
115.

2. R. B. Dean and W. J. Dixon (1951). ’Simplified Statistics for Small Numbers of
Observations', A/mlyiical Chemistry, 23, 636,

3. D. R. Read (1951). ‘Statistical Methods with Special Reference to Analytical
Chemistry’, Royal Institute of Chemistry. Lectures, Mnnnyraphs and Reports, No, I.

4. I. M. Kolthoffand P. J. Hiving (Ed.) (1950). Treatise on Analytical Chemistry. Part I.
Theory and Practice. Vol. I . Chapter 2. Errors in Chemical A naiysis. Ch. 3. Accuracy
and Precision, New York; Intcrsciencc Publisliers.

5. C. R. N. Strouts. J. M. Gilfillan and H. N. Wilson (1955). Analytical Chemistry. The
Working Tools. Volume II. Chapter 28. The Application of Statistical Methods to
Chemical Analy.sis, O.xford; Oxford University Press.

6. C. W. Wilson and D. W. Wilson (Ed.) (1959). Comprehenshe Analytical Chemistry.
Vol. I A. Classical .'inaly.us. Ch. 4. Statistics. Amsterdam; Elsevier Publishing
Company.

7. R, A. Fisher and F. ^'ates ( 1953). .Statistical Tahirs for Biological, Agriadtural and
Medical Research. 4th cdn., Edinburgh ; Oliver and Boyd.

8. D. J. Finney (1953). Experimental Design and its Statistical Basis. Cambridge;
Cambridge University Press.

9. C. A. Bennett and N. I., Franklin (1954). Statistical Analy.sis in Chemistry and the
Chemical Industry. New York ; John Wiley.

10. W. J. Dixon and F. J. Massey, Jr. ( 1957). Introduetion to Statistical Analysis. 2nd cdn.
New York; McGraw-Hill Book Co.

11. O. L. Davie.s (1957). Statistical Methods in Research and Production, seith Special
Reference to the Chemical Industry. 3rd cdn. Edinburgh : Oliver and Boyd.

12. H. A. Strobe! (1960). Chemical Instrumentation. A Systematic Approach to In-
strumental Analysis. Ch. 2. Errors of Measurement. Reading, Ma.xs. ; Addi.son-Weslcy
Publishing Co.

13. D. A. Pantony (1961), A Chemist's Introduction to Statistics. Theory of Error, and
Design of Experiment. Lecture Series. No, 2. Royal Institute of Chemistry. I-ondon.

14. J. Mandcl and F. J. Linnig (1956, 1958). ’Statistical Methods in Chemistry’,
Analytical Chemistry, 28, 770; 30, 739.

15. B. N. Nilson (1960). ’Statistical Methods in Chemistry’, Anaivtical Chemistrr, 32,
161R.

16. J. D. Hinchcn (1969). Practical Siaii.stirs for Chemical Research. London; Methuen
and Co.

17. M. J. Moroncy (1965). Facts from Figure.s. 3td cdn., revised. Penguin Books,
Hatmondsworth.

18. ILL. Youmans (1973). Statistics for Chemtsirv. Columbus, Ohio: Merrill Publishing
Co.

134

CHAPTER V SAMPLING

V, 1. THE BASIS OF SAMPLING. The purpose of analysis is to determine
the quality or composition of a material ; and for the analytical results obtained
to have any validity or meaning it is essential that adequate sampling procedures
be adopted. Sampling is the process of extracting from a large quantity of
material a small portion which is truly representative of the composition of the
whole material.

Sampling methods fall into three main groups :

1. those in which all the material is examined ;

2. casual sampling on an ad hoc basis;

3. methods in which portions of the material are selected based upon statistical

probabilities.

Procedure 1 is normally impracticable, as the majority of methods employed
are destructive, and in any case the amount of material to be examined is
frequently excessive. Even for a sample of manageable size the analysis would be
very time consuming, require large quantities of reagents, and would monopolise
instruments for long periods.

Sampling according to 2 is totally unscientific and can lead to decisions being
taken on inadequate information. In this case, as the taking of samples is entirely
casual, any true form of analytical control or supervision is impossible.

For these reasons the only reliable basis for sampling must be a mathematical
one using statistical probabilities. This means that although not every item or
every part of the sample is analysed the limitations of the selection are carefully
calculated and known in advance. Having calculated the degree of acceptable
nsk or margin of variation, the sampling plan is then chosen that will give the
maximum information and control that is compatible with a rapid turn over of
samples. For this reason, in the case of sampling from batches the selection of
individual samples is carried out according to special random tables (Ref. 1)
which ensure that personal factors do not influence the choice. -

V, 2. SAMPLING AND PHYSICAL STATE. Many of the problems
during sampling arise from the physical nature of the materials to be
s udied (Ref. 2). Although gases and liquids can, and do, present difficulties the
greatest problems of adequate sampling undoubtedly arise with solids.

1- Gases. Few problems arise over homogeneity of gas mixtures where the
s orage vessel is not subjected to temperature or pressure variations. Difficulties
may arise if precautions are not taken to clear valves, taps and connecting lines of

135

V, 2 QUANTITATIVE INORGANIC ANALYSIS

any otlicr gas prior to passage of the sample. Similarly care must be taken that no
gaseous components will react with the sampling and anal 3 'lical devices.

2. IJqiikls. In most eases general stirring or mi.xing is sufficient to ensure
homogeneity prior to sampling. Where separate piiases exist it is nccessar>’ to
determine the relative volumes of each phase in order to compare correctly the
composition of one phase with the other. The phasc.s should in any ease be
individually sampled as it is not possible to obtain a representative sample of the
combined materials even aficr vigorou.sly shaking the separate phases together.

3. Solids. It is with solids that real difficulties Over homogeneity arise.
Even materials that superficially have every appearance of being homogeneous in
fact may have localised concentrations of impurities and vary in composition.
The procedure adopted to obtain as representative a sample as possible will
depend greatly upon the tyjK- of solid. This process is of great importance since, if
it is not satisfactorily done, the labour and lime spent in making a careful analysis
of the sample may be completely wasted. If the material is more or less
homogeneous, sampling is comparatively simple. If. hosvcvcr, the material is
bulky and heterogeneous, sampling must be carried out with great care, and the
method will vary somewhat with the nature of the bulk solid.

The underlying principle of lire sampling of materia! in bulk, say. of a truck
load of coal or iron ore, is to select a large number of portions in a systematic
manner from difierent parts of the bulk and then to combine them. This large
sample of the total weight is crushed mechanically, if necessary, and then
shovelled into a conical pile. Every shovelful must fail upon ihcapc.x of tliecone
and the operator must walk around the cone ;is he shovels; this ensures a
comparatively even distribution. The top of the cone is then flattened out and
divided into quarters. Opposite quarters of the pile arc then removed, mixed to
foniKi .smaller conical pile, and again quartered. This process is repeated, further
crushing being carried out if nccc.ssary, until a sample of suitable weight (say,
200-300g) is obtained.

If the quantity of material is of the order of 2 3 kilos or less, intermixing may
be accomplished by the method known as ‘tabling’. The finely divided materia! is
spread on the centre of a large sheet of oilcloth or similar material. Each comer is
pulled in succession over its diagonal partner, ilie lifting being reduced to a
minimum: the particles arc thus caused to roll over and over on themselves, and
the tower portions are constantly brought to tlie top of the ma.ss and thorough
intermixing ensues. The sample may then be rolled to the centre of the cloth,
spread out, and quartered as before. The process is repeated until a sufficiently
small sample is obtained. The final sample for tlic laboratory, which is usually
between 25 and 200g m weight, is placed in an air-tiglit bottle. This method
produces what is known as the 'average sample’ and any analysis on it should
always be compared witli those of a second sample of the same material obtained
by the identical routine.

Mechanical methods also exist for dividing up particulate material into
suitably sized samples. Samples obtained by these means arc usually repre-
sentative of the bulk material within limits of less than ± 1 percent, and arc based
upon the requirements established by the British Standards Institution. Sample
dividers* exist with capacities of up to lOdm^ and operate either by means of a

• Av.nil.nblc from A. Galicnkump and Co Lut, P.O. Box 290, Tcclinico Itousc, Cliristopher St,
London, EC2R 2ER; Glen Creston. 37 The Broadway, Slanmnrc, Middlesex, and The Pascall
Engineering Co Ltd, Gatwick Road. Crawley, Sussex. RHIO 2 RS.

136

SAMPLING V,3

series of rapidly rotating sample jars under the outlet of a loading funnel, or by a
rotary cascade from which the samples are fed into a series of separate
compartments. Sample dividers can lead to a great deal of time saving in
laboratories dealing with bulk quantities of powders or minerals.

The sampling of metals and alloys may be effected by drilling holes through a
representative ingot at selected points ; all the material from the holes is collected,
mixed, and a sample of suitable size used for analysis. Turnings or scrapings from
the outside are not suitable as these frequently possess superficial impurities from
the castings or moulds. ,

In some instances in which grinding presents problems it is possible to obtain a
suitable homogeneous sample by dissolving a portion of the material in an
appropriate solvent.

Before analysis the representative solid sample is usually dried at 105-1 10 °C,
or at some higher specified temperature if necessary, to constant weight. The
results of the analysis are then reported on the ‘dry’ basis, viz., on a material dried
at a specified temperature. The loss in weight on drying may be determined, and
the results may be reported, if desired, on the original ‘moist’ basis ; these figures
will only possess real significance if the material is not appreciably hygroscopic
and no chemical changes, other than the loss of water, take place on drying.

In a course of systematic quantitative analysis, such as that with which we are
chiefly concerned in the present book, the unknowns supplied for analysis are
usually portions of carefully analysed samples which have been finely ground
until uniform.

It should be borne in mind that although it is possible to generalise on
sampling procedures all industries have their own established methods for
obtaining a record of the quantity and/or quality of their products. The sampling
procedures for tobacco leaves will obviously differ from those used for bales of
cotton or for coal. But although the types of samples differ considerably the
actual analytical methods later used are of general application.

V, 3. CRUSHING AND GRINDING. If the material is hard (e.g., a sample
of rock), it is first broken into small pieces on a hard steel plate with a hardened
hammer. The loss of fragments is prevented by covering the plate with a steel
ring, or in some other manner. The small lumps may be broken in a ‘percussion’
mortar (also known as a ‘diamond’ mortar) (Fig. V, 1 ). The mortar and pestle are
constructed entirely of hard tool steel. One or two small
pieces are placed in the mortar, and the pestle inserted into
position ; the latter is struck lightly with a hammer until the
pieces have been reduced to a coarse powder. The whole of
the hard substance may be treated in this manner. The coarse
powder is then ground in an agate mortar in small quantities
at a time. A mortar of mullite is claimed to be superior to one
of agate; mullite is a homogeneous ceramic material that is
harder, more resistant to abrasion and less porous than
agate. A synthetic sapphire mortar and pestle (composed
. . essentially of a specially prepared form of pure aluminium

03ude) IS marketed; it is extremely hard (comparable with tungsten carbide) and
will grind materials not readily reduced in ceramic or metal mortars. Mechanical
(motor-driven) mortars are available commercially.

137

V, 4/5/6 QUANTITATIVE INORGANIC ANALYSIS

V, 4, HAZARDS IN SAMPLING. The handling of many materials is
fraught with liazards (Ref. 3) and this is no less so wlicn sampling materials in
preparation for chemical analysis. The sampler must always wear adequate
protective clothing and if possible have detailed prior knowledge of the material
being sampled. When dangers from to.xicily exist the necessary antidotc.s and
treatment procedures should be available and established before sampling
commences (Ref. 4). In no in.stanccs should naked flames be allowed anywhere
near the .sampling area.

Apart from the toxic nature of many gases the additional hazards arc those of
excessive release of gas due to pressure changes, spontaneous ignition of
inflammable gases itnd sudden vaporisation of liquified gases.

With liquids dangers frequently arise from easily volatiscd and readily
inflammable liquids. In all eases precautions should be greater than under
normal circumstances due to the unpredictable nature and conditions of taking
samples. The sampler must always be prepared for the unexpected, as can arise,
for example, if a container has built up excess pressure, or if the wrong liquid has
been packed. Toxic and unknown liquids should never be sucked along tubes or
into pipettes by mouth.

Even the sampling of solids must not he casually undertaken, and the operator
should always use a face mask as a protection until it is established that the
powdered materia! is not haz.irdous.

It should he borne in mind that sampling of radioactive substances is a
specialist opcnition at all times and .should only be carried out under strictly
controlled conditions within restricted areas. In almost ;dl instances the operator
must be protected against the radioactive emanations from the substance he is
sampling.

Correct sampling of materials is therefore of importance in two main respects;
first to obtain a representative portion of the material for analysis, and secondly
to prevent the occurrence of accidents when sampling hazardous materials.

V, 5. References

1. J. Murdoch and J. A. Hrirncs (I‘)70). Siaiittiral HiHcs for Sik'iicc. Eitiiinccriu!; ami
Mumii’fniciu. 2nd cdn. London. Macmillan, pp. 3(t .t3.

2. C. R. N. Stroms, J. H. Gdfillan and IL N. WiLon (1955,1. Analytical Chrmisiry. The
Warkint; Tonis. Vol. 1. Ch. .3, Samplmy. London ; Oxford University Press.

3. N, Irving Sax (Ld.) (1968). Dan^crou', I'lopiriics of Industrial ,({iiicriah. 3rd cdn.
Rcinhokl, New York,

4. G. D. Muir, (Ed ) (1971). llazunh in the Chemical Lahnrawry. London: Royal
Institute of Chemistry.

V, 6. Sclecicd bibliography

1. C. L. Wilson and D. W. Wilson ( Ed.) { 1959.). Comprchcit.stYc .Analytical Chemistry. Vol.
lA, Classical Analy.si.s, Ch. 11.3., Samplin};. Amsierdam; Elsevier Publishing Co.

2. H. A. Lairinen (1960). Chemical Anahsis. .An Adeaneed Text and Reference. Ch. 27.
Sampling. Nevy York; McGraw-Hill BookCo.

3. N. V. Sleere(Ld.)(1967). Handbnak of l^ihoratorv Safety. Cleveland. Ohio; Chemical

Rubber Co. ’ ...

138

PARTC SEPARATIVE TECHNIQUES

139

PARTC INTRODUCTION

An ideal analytical method would enable a species to be determined directly in
various matrices. Few, if any, analytical measurements are, however, wholly
specific for a single species so that a major problem in quantitative analysis is the
elimination of interferences. The following two general methods are available for
dealing with substances that interfere in an analytical measurement :
determination of uranium as 8-hydroxyquinolate.

1. Masking of the potential interference(s) so as to prevent it contributing to
the measurement step (Ref. 1 ) ; this is commonly effected by the introduction of a
complexing agent that reacts selectively with the interfering substance, e.g., the
masking of iron(III) and aluminium by EDTA in the solvent extraction and

2. The isolation of the species to be determined in a separate phase from the
interfering species, by means of one of the various separative techniques, e.g., ion
exchange, gas-liquid and liquid-liquid chromatography, solvent extraction.
Such separations are, of course, based upon equilibrium processes so that
complete separation of the interference from the species required is never
possible. In practice the aim of the separation procedure will be to lower the
concentration of the interference to a tolerable level while at the same time
ensuring that any losses of the desired constituent are smaller than the allowable
error in the analysis. The level of interfering substance which can be tolerated will
be dependent upon the relative sensitivity of the final analytical measurement for
tte interfering and required constituents. If the measurement is much less
affected by the interfering substance (i.e. lower sensitivity), then a partial
separation may be adequate, but if the sensitivity of the method for the two
constituents is about the same then a virtually complete separation may be
required.

In trace analysis, where the ratio of required minor constituent to major
component may be as small as 10“® or 10 the separation procedure may also
effect a useful preconcentration of the trace constituent, thus providing an
adequate amount of the substance for the measurement to be employed.

The aim of the present section is to survey and illustrate the application of the
chiet separative techniques in inorganic quantitative analysis. Wherever possible

e use of an appropriate instrumental method is indicated for the final analytical
m^surement, as well as titrimetric or other classical methods.

0 attempt has been made to deal with the general theory of separatory

141

INTRODUCTION

methods since this is already adequately considered in various analytical
chemistry texts (Refs. 2, 3 and 4).

References

1. D. D. Perrin (1975). ‘Selection of Masking Agents for Use in Analytical Chemistry.’
CRC crit. Ri'y. analyi. Chem. 5. 85.

2. L. B. Rogers (1961), Principles of Separations’, in Treatise on Analytical Chemistry.
Part I (cd. i. M. Kolthoffand P . }. Elring). Vol. 2, New York; Inicrscicncc.

3. B. L. Kargcr. L. R. Snyder and C. Horvath (1973). An introduction to Separation
Science. New York; Wiley.

4. J. A. Dean (1969). Chemical Separation Methods. New York; Van Nostrand
Rcinhold Co.

142

CHAPTER VI SOLVENT EXTRACTION

VI, 1. GENERAL DISCUSSION. Liquid-liquid extraction is a technique in
which a solution (usually aqueous) is brought into contact with a second solveriT
(usually organic)7essentially immiscible, with the first, in order to bring about a

transfer*5r 5 n^r more solutes into the second solvent.l The separations that can

Tin many cases separation

be performed are^mple, clean, rapid, and convenient.

may be effected by taking in a separatory funnel for a few. minutes. The
technique is equally applicable to trace level and large amounts of materials.* We
are concerned largely with samples in aqueous solution, and the production of
chelate and of ion association extraction systems.

To understand the fundamental principles of extraction, the various terms used
for expressing the effectiveness of a separation must first be considered. For a
. y^tite A distributed between two immiscible phases a and b, the Nernst
ji^nbution (or partitionl law states that, provided its molecular state is the sanie

both liquid s and that the temperature is<;6nstant: ‘~^ I

^ieficentration of solute in solvent a
where Kp is a constant known as the distribution (or partition) coefficient. The

as stated, is not thermodynamically rigorous (e.g., it takes no account of the
activities of the various species, and for this reason would be expected to apply
only in very dilute solutions, where the ratio of the activities approaches uhity),
but IS a useful approximatio nniie law in its simple form does not apply when the
.Slgili ^ting species undergo^dissociation or association in either phase.^ n the
practical applications ofsolvent extraction we are interested primarily 4n 'the
raction of the total s olute in one or other pha se, quite regardless of its mode of
issociation, association, or interaction with other dissolved species. It is
convenient to introduce the ter m distribution ratio D (or extraction coeffi cient E):

»=(Si/(c.))V ' ^ ~

"'h^reihesymSoR^^ denotes the concentration of .4 in all its forms as determined
analytically;

varll**’® examples in this chapter, other (and more complex) examples will be found in

s-napter XVin, e.g., Section XVm, 28.

143

VI, 1 QUANTITATIVE INORGANIC ANALYSIS

A problem often encountered in practice is to determine what is the most
efficient mctliod for removing a substance quantitatively from solution. It can be
shown that if V cm^ of, say, an aqueous solution containing .Vo g of a solute be
extracted /) times with r-cm^ portions of a given solvent, then the weight of solute
.v-„ rem ainine in the water layer is giv en by the expression : ^ i

-7 (m^J o,. 'or.

v-d'

where D is the distribution ratio h elw>u:a. walcr and_lhe given solven t. It folio ws^
therefore, that the best method of extraction with a given volume of extracting >
liquid is to employ several fractions of the liquid rather than to utilise the whole
quantity in a single extraction. to~ q

Let us take a particular extimple. Let us suppose that we shake 50 cm^ of water ^
containing O.I g of iodine with 25 cut' of carbon tetrachloride. Thc'distribuiion '•
coefficient of iodine between water and carbon tetrachloride at the ordinary^
laboratory temperature is l/85.i.c.,at equilibrium the iodine concentration in the
aqueous layer is l/85th of that in the carbon tetrachloride layer. We will compute
the weight of iodine remaining in the aqueous layer after one extraction with 25-
cm^ and also after, say. three extractions with cm-’ of the solvent by
application of the above formula. The former can be simply computed as follows.
If.v, g of iodine remains in the 50 cm-’ ofwater, its concentration i.<3q/50 gcm"’;
the concentration in the carbon tetrachloride layer will be (0.1 g cm"’.

Hence:

(0.1 - .v7)725

. or .V

0.00230 g

The concentration in the aqueous layer after three extractions with 8.33 cm’ of
carbon tetrachloride is given by;

-’■ 3 .=

0.1

(1/85) X 50 y
(50/85) + 8 . 33 )

0.0000145 g

The extraction may therefore be regarded as virtually complete.

If we confine our attention to thcLlljstribinjpn.pJ.a solu te A between water and
an organic sol vent, w-c may w rite th e percentage extraction E. as;

— rO017r),l; UK)/a ~

wher e I-^, and7h....J CCPresarMhc-xutku nes~ of the organic and aqucou sjhases
respectively. Thus the pcrccntagc^fcxtraction varies wath the volume ratioTifThe

two phases and t he disTributimi coenicieni. =rj> "

/Tfl he solution contains two solutes A and /I it often ha p pens that under th e
conditibn.s favouring the complete c,xtraciion ofA, some B is extracted as well.
J[lic.£irc£liyenc ss of separation increases with the magnitude of the scparation_;
.cocfficjc nt or iaHor /I, w iTtclrfe~TClrtled nrThtriTtdTYidOaTdTstribuiion ratios as
follows; . .. .

144

v'-pv'-t

SOLVENT EXTRACTION VI, 2

_ lOandZJfi = 0.1, a single extraction will remove 90.9 per cent of v4 and 9.1
per cent of B (ratio 10:1); a second extraction of the same aqueous phase will
bring the total amount of A extracted up to 99.2 per cent, but increases that of B to
17.4 per cent (ratio 5.7:1). More complete extraction of A thus involves an
increased contamination by B. Clearly, when one of the distribution ratios is
V relatively large and the other very small, almost complete separation can be
Iquickly and easily achieved. If thC' separation factor is large but the smaller
distribution ratio is of sufficient magnitude that extraction of both components
occurs, it is necessary to resort to special techniques to suppress the extraction of
the unwanted component.

VI, 2. Actors favouring solvent extraction, it is well

known tha t hydrated inorganic sa|ts tend to be more soluble in water than, in
organic solvents such as benzene, chlor oform; etc., whereas organic subs tances
Tend to be more soluble in organic solvents' than in'water un less they incorporate"
a sufficient numb er of hydroxyl, sulphonic. or other hydrophilic groupings. In
solvent extracti^ analysis tor metals we are concerned with methods.by which
the water solubility of inorganic cations may be masked by interaction with
appropriate (largely organic) reagents ; this will in effect remove some or all of the
water molecules associated with the metal ion to which the water solubility is due.

Ionic compounds would not be expected to extract into organic solvents from
aqueous solution because of the large loss in electrostatic solvation energy which
would occur. The most obvious way to make an a queous ionic species extractable
|s to neutralise its charge. This can be done by formation of a neutral metat
l ^ate complex or. by ion, association :' ine larger and ~more hydrophobic the
renting molecular species the better WilJ belts extraction.

' jn chelation complexes (sometimes called inner complexes when uncharged)
thFcentral metal ion co-ordinates with a polyfunctional organic base to form a
stab le p ng compo und, e.g.. copperdll ‘acetvlacetonate’ or irondID tcupierrate’:

CHj— C=0 CH3— C— 0^1^

CH3— c=0 CH3— c=o

CH3— c Cu/2

HC O

C— CH,

N=0

iFe"

The f

cussed in's°^^‘ influence the stability of metal ion complexes are dis-
ection X, 20, but it is appropriate to emphasise here the significance

145

VI, 3 QUANTITATiVK INORGANIC ANALYSIS

or the chelate effect and to list the features of the ligand which affect chelate
formations

(i) Tiff basic streiu’th (if tlw {•hclati/1,1; group. The stability of the chelate
c^jdexes formed by a given metal ion generall y increase with increasina basic
strength ;Pf the chelating agent, ^ sjrifiasut£d by the values —

. nature oj the tionor atoms JnJ ht:.xhdami^-a£iU^ Ligands which ,

contain donor atoms of th e soft base type form their most stable complexes wit h
the relatively sm all group ^class (bl metal ion s (i.e., soft acids) and arc thus more
selective reagents, dhis is illustrated by tTtc reagent diphcn^tfiiocarSaTOni;

' (dithigone)iire'd for thesolventcxirac lion of metal ions stKJiasRd- \Ag\ Hg^
Cu^^-^Bi-'^.Pb*’ and Zn~'! ~ — -*■

(id) RiitK size. Five - or six-membered conjugated chelate rings are most
^able since these liaveTninimum strain. The functional groups of the ligand must
’Ti^arsmTated that tHcy'pcrmil the forltiation of a stable ring.

(iy { Resonanre ami stcric clTi'cts, The .stability of chelate structures is
cnimneed by contributions of resonance structures of the chelate ring; thus
jc;opper acel yl acetonatc (see formula above) has greater stability than the copp er
chelate of salicylald o.ximc, A good c.xamplc of stcric hindrance is 2 . 9 -
'’oimeihviDiicnilhihrolinc (neociipr oincl. which does not gnT a c rin Ttr lex w ith-
!ron{ll)asd^csilie-«rTstibsiuutctrpIIg^nthroiinc; this hindrance is at a minimum
“in me teiFahcdral grouping of the reagcnriTTOlcculcs about a univalent tetra-
coordinated ion such as that of coppcifl). A nearly specific re.agcnt for copper is
thus available.

nie choice of a .satisfactory chelating agent for a particular separation should,
of course, lake all the above factors into account. The critical infiuence of pH on
the solvent extraction of incial chelates is .shown in the following section.

^{, 3 . QUAiNUTATIVE treatment of solvent extraction

EQUILIBRIA. The solvent extraction of a neutral metal chelate cgmplo f
, formed from the chclalingapen L l iR.a ccording to the equation (

M"^ -t-nR-

may be treated quantitatively on the basis of t he follovdncjissiimpiions: (a) that
th e reagent and the metal complex exist assimptinTiTassociated molecules in both
phases ; (/)).solvatioirplays no significant part in th e exirn. ciiaxmtgcc$s; and (c) llie,
solutes arc uncnTi r g ^etijtuokculesjio d ttic ir concent^ions arc generally so loj^w

that the behaviour of their solutions dcpaFiirtirrl c fi 0111 identity. 1 he (.iis.soci.i ii^
of the chelating agent HR in the aqueous nh ase is rcpre.scnted by the equation

HR:;±ir-FR-

Thc various equilibria involved in the solvent extraction process are cxprc.sscd in
terms of the following thermodynamic constants:

/ dissociation constant of compic,x. A', = IM'’'*],..[R‘]:./IMR„]„.

Dissociation constant of reagent. A, = IH'’]„,(R'']„./[HR];,.‘--^

1 Partition coefficient of complex, p, = (M R /{MR„]„

Partition coefficient of reagent, p, - }HR]„./[HR]„

where the subscriptsjii ind r ref er to complex and reagent, anc Liv and o to aqueo us
aad.organieplio.se rcspecnTCly. ~ ~

Tlie distributiorTratibrlrc.. ihc_j-atio of the amount of metal extracted as

146

SOLVENT EXTRACTION VI, 3

; ^ai^^ase

complex into the org;

^se, is given by' .

[MR„L/{[MR„],,+[M"+]„}

which can be shdwn (Ref. 1) to reduce to

.where [K^p^flK^p^

If the reagent concentration reihains virtually constant

to that remahrifig in all forms in the aqueous

■y

D = K*I[H^Z where_^;;;^£[HR|__

and the percentage of solute extracted, E, is given by

Ipg£-Iog(100— £) = logD

= logK* + npH

— Zoo, Qw t)

1 '

io a K,* — .to Ip-j *2

The distribution of the m etal in a given system is a function of the p H alone. The
equation rejiresents a tamity ot sigmoid curves when E is plotted against pH, with
the position of each along the pH axis depending only on the magnitude of K* ’
and the slope of ea'ch'^niquely depending upon n. Some theoretical extraction
curves for divalent metals showing how the position of the curves depends upon
the magnitude of X* are depicted in Fig. VI, 1; Fig. VI, 2, illustrates how the slope -
depends upon n. It is evident that a ten-fold change in reagent concentration is
exactly offset by a ten-fold change in hydrogen-ion concentration, i .e.. by a
change of a single unit of pH: such a change of pH is much easier to effect jn
pfactice.JfpHi is defined~aFthe pH value at 50 per cent extraction (£./ = 50) we ^

see from the above equation that

5i ^

H}= —logK*

pHt values of two metal ions in a specific system is a measure of
^^^eseparataEni!y_^Tl^two ions. If the pHi values are siifflcientlV far apaiT.7thdP~

^^^‘^^^^ratidiL caiL-be a^ chieved by conTrolling ihe 'pH^Q^gjcSSCtionrR^fr-
oroti helpful to plot thewtractionTcurves ol metal chelates. If one takes as the-
cnterion of a successful single-stage separation of two metals by pH control a 99
per cent extraction of one with a maximum of 1 per cent extraction of the other,

147

VI, 4 QUANTITAWE INORGANIC ANALYSIS

fnr hivnlcnl metals a difTerencc of two pH units would be ncccssiiry betwe en the
' two piyv^es; the diircrence is jess lor tcrvalent metals. So me figures for The
extraction of metal dithizon;Ucs in chloroform arc: ”^

Metal Ion
Optimum pH of
Extraction

Cu(U)

Hrdi)

1-2

Sn(II)

6-9

7-9

Ni,Zn Pb

S 8.5-11

If the pH is controlled by a buffer solution, then those metals with pHj values in
this region, together with all metals having smaller pHj values, will be extracted.
The nHi values may be altered (and the selectivity of the extraction thus
mcreasSj jtn’ the use of a competitive complexin g ti geni or of masking agenLsy
Thus in thVseparation of mercury and copper TTv extraction with diTluzone inyv

-( carbonTtciracliTor idc at pH 2 thcTiddiiioii oi t:DT'A torms a walcr-soIuF

complex whicii completely TuasKThe Coppcrtful does not allect the mefctTfy

extraction. Cyanides rais e tliiFpl li valu es ol mercury, copper, zinc, an Jeadnuum —
m dithizone extraction with carbon tctraHilbridci

4. 10N-A.SS0CIAT10N COMPLKXIiS. \ln ibn-associalion'.bompicxcs"
the i norganic ion a ssociates with oppositely charec d ions to fo nn_a- nculral <^‘'
extractable species. Sucji_cgmplcxcs niaT' ionn ^usler.s with in'cfeasing~‘t

1 '-

coiiceniration which arc larger than just simple ion pairs, particularly in organic (
solvents of Tow dielectric constant. The Rtllowing types of ion-association
^ c^ploxes may be rccogni.sed :

hose formed from a reagent viclding~u~1 .TrgcrpTK(iiiicJon’. c.g.. the
tetruphcnvlarsonium. TC TTf %As ‘ . antrierrnfartTlnmTTrominTirTff^ClTlQl .sN'*'
tons, which fo rm large ion aeercaates or clusters xnt h suitable o jiiiosiicly chared

These la r ge and hiilkv luiis do noi Inrvira

ions, ergr.Tlie pe rrlicnatc ion. Rc0.i _ _

priluaiy hydration siiell and cau^ disruption of the hydrogen-bonded water
' stru cture; the larger th e ion the a rcatcr the amount ol disruption and the grcatTr
thcTciidcncy lor the ion associatiotrspecies to be otl-nTeCt int o me organic riliaS :,
'^Hlesc large ion extraction sy.sTcms lack specificity since any relativclyjarge
^‘unhydrated univalent cation will extract any such large univalent anion. On The
, othcrHancr pblvvalcnt ionsTT fccIuTsirorrlterr ereaier ludraiion cnergv.lTrc not so
, S easily extracted and good~sc~parat'ionx are possible between MnOJ^. RcOj" or
* ifnd CTU 4 ^^ \ MoOT^' . for example. ~~

^i jbrt-lrostrfnvoTxin.g a caiiomgTTnnnTmfrT^ of a metal ion. .Thus

clwlfting*agcnts liaving tvro tuKdiaf^eddonoTlitdm^ sFcH as TnO-p henan:
throlinc, form catitmicchelatc complexes which are large and hydrocarbon liliei
/TT tgTbheiiTihniroTi n cTTronl 1 1 ) pcrcldoraie extracts fairly well into chloroform.
an d_cxtract!.on is virtuiTnv comple te usmgTargeliTubns such as long cluitfralRyl

Dagnall and West {Rel. TnuTve described the formation and extraction of a
blue reniciry complex, AgXlLl.lO-ph enanihroline-Bromopyrogallol Red (BPR).
as the basis of a highly sensitive spetdr^imo'metne pTw^ for the
dclermination of traces of silver (Section VI, 15). Tlie reaction mechanism for the
formation of the blue complex in aqueous solution was investigated by
photometric and potentioractric methods and these studies led to the conclusion
that the complex is an ion-association system. (Ag{'phcn);)jBPR- ", i.c.. involving
a cationic chelate complex of a metal ion (Ag"') associated with an anionic

148

SOLVENT EXTRACTION VI, 5

counter-ion derived from, the dyestuff (BPR). Ternary complexes have been
reviewed by Babko (Ref. 8).

Types (i) and (ii) represent extraction systems involving coordinately
unsofeated large ions and differ in this important respect from type (iii).
(niiljtlios ein which solvent molecules are directly involved in formation of the

lOn-association complex. Most of the solve nts (ethers, e sters, ketoijes and"
&olsr ^ch participate in. this way contain donor oxygen atom^i(iia~Tlre
coordinating ability of the solvent is of vital significance. T he coordinate solvent
molecules facilitate the solvent extraction of salts such as chlorides and nitrates
by contributing both to the size of the cation and the resemblance of the complex
to^^olvent. ~ ~ ^

'^A class of solvents which shows very marked'sblvating properties for inorganic /
compounds comprises t he esters of orthophosphoric acid. Th e functional group
i n these molecules is the semipolar phosphoryl group. — 0~, which has a

basiT oxygen atom with good steric availability . A typicaT compound is tri-n-^
b^utyl phosphate (TBP) .^ich has been widely used in solvent extraction on both
the laboratory and industrial scale; of particular note is the use of TBP for the
extraction of uranyl nitrate and its separation from fission products.

The mode of extraction in these ‘oxonium’ systems may be illustrated by
considering . the ether extraction of iron(III) from strong hydrochloric acid
solution. In the .aqueous phase chloride ions replace the water molecules
coordmated to the Fe^"^ ion, yielding the tetrahedral FeCU~ ion. It is recognised
that the h ydrated hydronium ion, H^0'*~(H, 0), o r HpO^^, normally pairs with
t he complex halo^anibns, put in the preience of the organic solvent; solvenf
molecules enter the aqufaus pha se And compet e with water for positions in th e
.MSPibn shell ol the protorf. On this basis the primary species extracted into the

^her (R ^O) phase is considered to be [H 30 (Rj 0 ),'^,FeCl 4 1 although
legation oiThts^yecies luuy occui in sol vents of low dielectric conkant.

5. EXTRACTION REAGENTS. Many complexes of metals in aqueous
solution are coloured: when extracted with an organic solvent, the coloured
extract may be used directly for the determination of the concentration of the
yy ® ^°*orirnetricor, preferably, spectrophotometric techniques (see Chapter
II). These techniques are particularly applicable with many chelate
^ number of coloured inorganic complexes, such as
, 0 ybdenum blue’ and iodobismuthite ion, can be treated in this way. In this
c ion we shall discuss a limited number of chelating and extraction reagents,* as

as some organic solvents with special interest as to their selective extraction
properties.

aceti 4'®^^acetone (Pentane-2, 4-dione) CHsCO-CHa’COCHa. Acetyl-
water colourless mobile liquid, b.p. 139 °C, which is sparingly soluble in
useful h tii ^ cm^ at 25 °C) and miscible with many organic solvents. It is
etc 1 anH° ^ ^ solution (in carbon tetrachloride, chloroform, benzene, xylene,
chelates The compound is a /?-diketone and forms well-defined

soluble in'"* ” metals. Many of the chelates (acetylacetonates) are

unlike solvents, and the solubility is of the order of grams per dm^,

0 most analytically used chelates, so that macro- as well as micro-

‘ The formulae of the

compounds not given in the text will be found in Section XI, 11.

149

VI, 5 QUANTITATIVE INORGANIC ANALYSIS

scale separations arc possible. The selectivity can be increa.sed by using EDTA as
a masking agent. The use of acctylacctone as both solvent and extractant (c.g., for
Al,Be,Ce,Co(III),Ga.ln,Fc. U(VI),ctc.|oncrs.scveraladvantagcsovcritsuscin
solution in carbon tetrachloride, etc.: extraction may be carried out at a lower pH
than otherwise feasible because of the higher reagent concentration: and often the
solubility of the chelate is greater in acctylacctone than in many organic solvents.
The solvent generally used is carbon tetrachloride; the organic layer is heavier
than water.

An interesting application is the separation of cobalt and nickel: neither Co(I!)
nor Ni(II)form e.xtractablcchclatc.s. but Co(1 11) chelate is extractable; extraction
is therefore possible following oxidation.

2. Thcnoyltrifluoroiicctonc, (TTA), C 4 H 3 S-CO CH, COCF 3 . This is a
crystalline solid, m.p. 43 X; it is, ofcour.se, a ^-dikctonc. and the trifluoromcthyl
group increases the acidity of the cnol form so that extractions at low pH values
are feasible. The reacliviiy of 2TA is similar to that of acctylacctone; it is
generally used as a O.l-O.SAf solution in benzene or toluene. The dificrcncc in
extraction behaviour of hafnium and zirconium, and al.so among lanthanoids :ind
actinoids, is especially noteworthy.

3. 8-HydroxyqtiinoIine (oxine). Oxinc is a ver.salilc organic reagent and

forms chelates with many metallic ions. The divalent and trivalent metal chelates
have the general formulae and M(CoH60N)j: the oxinates of the

higher-valent metals may differ somewhat in composition, c.g., CefC^HsON)^;
Th(C,H,,ON)4 (CgH,ON); WO,(C„IUON)j : MoO;,(C!,H^,ON)j :
U 30 f,(C 9 Hf, 0 N)f,-(C 9 H 70 N). Oxine is generally used as a 1 per cent (0,07Af)
solution in chloroform, but concentrations as high as 10 per cent arc
advantageous in some eases (c.g., for strontium).

8-Hydroxyquinolinc. having both a pltcnolic hydroxyl group and a basic
nitrogen atom, is amphoteric in aqueous solution ; it is completely extracted from
aqueous solution by chloroform at pH < 5 and pH > 9; the distribution
coefficient of the neutral compound between chloroform and water is 720 at
18 °C. The usefulness of this sensitive reagent has been extended by the use
of masking agents (cyanide, I: DTA, citrate, tartrate, etc.) and by control of pH.

4. Dimcthylglyoximc. The complc,xcs with nickel and with palladium are
soluble in chloroform. The optimum pH range for extraction of the nickel
complex is 4-12 in the presence of tartrate and 7-12 in the presence of citrate
(solubility 35-50 /rg Ni per cm ’ at room temperature): if the amount of cobalt
exceeds 5 mg some cobalt may be extracted from alkaline solution. PalladiumfH)
may be extracted out ofeu. 1 A/-sulphuricacid solution.

5. I-Nitroso-2-nnplithoI. The reagent forms extractable complexes
(chloroform) with Coflll) in an acid medium and with Fc(II) in a basic medium.

6. Cupferron (ammonium .salt of N-nitroso-N-phcnylhydroxylaminc). The
reagent is used in cold aqueous solution (about 6 per cent). Metal cupferrates are
soluble in diethyl ether and in chloroform, and so the reagent finds wide
application in solvent-extraction separation schemes. Thus Fe(IlI), Ti, and Cu
may be extracted from 1.23/ HCl solution by chloroform: numerous other
elements may be extracted largely in acidic solution.

7. Diphonylthiocarbazonc (difhizone), CfiH 5 N=N CS NH NH C 6 H 5 .
The compound is insoluble in water and dilute mineral acids, and is readily
soluble in dilute aqueous ammonia. It is used in dilute solution in chloroform or
carbon tetrachloride. Dithizone is an important sclectivereagent for quantitative

150

SOLVENT EXTRACTION VI, 6

determinations of -metals: colorimetric (and, of course, spectrophotometric)
analyses are based upon the intense green colour of the reagent and the
contrasting colours of the metal dithizonates in organic solvents. The selectivity
isimproved by the control of pH and the use of masking agents, such as cyaiiide,
thiocyanate, thiosulphate, and EDTA.

i Sodium diethyidithiocarbamate, {(C2H5)2N-CS S}~Na^. This reagent
is generally used as, a 2 per cent aqueous solution; it decomposes rapidly in
solutions of low pH. It is an effective extraction reagent for over twenty metals
into various . organic solvents, such as chloroform, carbon tetrachloride, and
ethanol. The selectivity is enhanced by the control of pH and the addition of
masking agents.

9 . Toluene- 3 , 4 -dithiol (dithiol), CH3-C6H3-(SH)2. This compound is a
solid, m.p. 31 °, which forms complexes in acid solution with Mo(VI), W(VI), and
Re{VI) that are extractable by chloroform or pentyl acetate. Control of the pH or
the use of citric acid permits the selective extraction of molybdenum.

10 . Tri-n-butyl phosphate, («-C4H9)3P04. This solvent is useful for the
extraction of metal thiocyanate complexes, of nitrates from nitric acid solution
(e.g., cerium, thallium, and uranium), ofchloridecomplexes, and ofaceticacid from
aqueous solution. In the analysis of steel, iron(III) may be removed as the soluble
‘iron(III) thiocyanate’. The solvent is non-volatile, non-inflammable, and rapid in
its action.

11 . Tri-n-octylphosphine oxide, (n-C8Hj7)3PO. This compound (TOPO)
dissolved in cyclohexane (O.IM) is an excellent extraction solvent. Thus the
distribution ratio of U(VI) is of the order of 10 ® times greater for TOPO than for
tri-«-butyl phosphate. The following elements are completely extracted from IM-
Jydrochloric acid: Cr(VI) as H2Cr207,2T0P0; Zr(IV) as ZrCl 4 , 2 TOPO; Ti(IV);
U(VI) as U02(N03)2,2T0P0; Fe(III); Mo(VI) and Sn(IV). If the hydrochloric
acid concentration is increased to 7 M, Sb(lII), Ga(III) and V(IV) are completely
extracted.

W, 6 . SOME PRACTICAL CONSIDERATIONS. Solvent extraction is
generally employed in analysis to separate a solute (or solutes) of interest from
su stances which interfere in the ultimate quantitative analysis of the material;
sometimes the interfering solutes are extracted selectively. Solvent extraction is
LTly ^ species which in aqueous solution is too dilute to be

tions-^ solvent for extraction is governed by the following considera-

W A high distribution ratio for the solute and a low distribution ratio for
undesirable impurities.

bow solubility in the aqueous phase.

u ciently low viscosity and sufficient density difference from the aqueous
fivi T uvoid the formation of emulsions.

IV) Low toxicity and inflammability.

ase of recovery of solute from the solvent for subsequent analytical

solvent and the ease of stripping by chemical
Somef ” ^ uierits attention when a choice is possible.

Salting-ni^^^ solvents may be used to improve the above properties.

Extr improve extractability.

■■ac ion. Extraction may be accomplished in either a batch operation or

151

VI, 6 QUANTITATIVE INORGANIC ANALYSIS

a continuous operation. Batch extraction, the .simplest and
most widely used method, is employed where a large
di-stribiition ratio for the desired .separation is readily
obtainable. A small number of batch extractions readily
remove the desired component completely. It may be carried
out in a simple separatory funnel. For solvents lighter than
water the modified separatory funnel (Fig. VI, 3), designed to
simplify the removal of the lighter phase, may be used. After
equilibration, the lighter (c.g.. ethereal) and aqueous layers arc
displaced upwards by introducing mercury through the
stopcock at the bottom of the bulb with the aid of a subsidiary
mercury levelling bulb. Thc.stopcocks should be well ground so
that a lubricant is not required: if a lubricant is used at all it
siiould preferably be a silicone-type grease.

The two layers arc shaken in ;i separatory funnel until
equilibrium is attained, after which the two layers arc allowed
to .settle completely before sampling. Tlie c.xtraction and
sampling should be performed at constant temperature, since
the distribution ratio as well as the volumes of tlie .solvent are influenced by
temperature changes. It must be borne in mind that too violent agitation of the
extraction mixture often serves no useful purpose: simple repeated inversions of
the vessel suflicc to give equilibrium in a relatively few inversions. If droplets of
aqueous phase are entrained in the organic extract it is possible to remove them
by filtering the extract through a dry filter paper: the filter paper should be
washed several times with fresh organic solvent.

When the distribution ratio is low continuous methods of extraction are u,sed.
This procedure makes use of a continuous flow of immiscible solvent through the
solution; if the solvent is volatile, it is recycled by distillation and condensation
and is dispersed in the aqueous phase by means of a sintered glass di.se or
equivalent device. Apparatus arc available for cflecting such continuous
extractions with automatic return of the volatilised .solvent (see Selected
Bibliography at the end of this chapter).

.Stripping. Stripping is the removal of the extracted solute from the organic
phase for further preparation for the detailed analysis. In many colorimetric
procedures involving an extraction procc.ss the concentration of the desired
solute is determined directly in the organic phase by measuring the absorbance of
a known volume of the solution of the coloured complex.

Where other methods of analysis arc to be employed, or where further
separation steps arc necessary, the solute must be removed from the organic
phase to a more suitable medium. If the organic solvent is volatile (c.g., diethyl
ether) the simplest procedure is to add a small volume of water and evaporate the
solvent on a water bath; care should be taken to avoid loss of a volatile solute
during the evaporation. Sometimes adjustment of the pH of the solution, change
in valence state, or the use of competitive water-soluble complexing reagents may
be employed to prevent loss of the solute. When the extracting solvent is non-
volatile the .solute is removed from the solvent by chemiad means, c.g., by shaking
the solvent with a volume of water containing acids or other reagents, whereby
the extractable complex is dcconipo.scd. The metal ions arc tiicn quantitatively
back-extracted into the aqueous phase.

Impurities present in the organic phase may sometimes be removed by

152

SOLVENT EXTRACTION VI, 7

backwashing. The organic extract when shaken with one or more small portions
of a fresh aqueous phase containing the optimum reagent concentration and of
correct pH will result in the redistribution of the impurities in favour of the
aqueous phase, since their distribution ratios are low; most of the desired element
will remain in the organic layer.

Completion of the analysis. Having separated a particular element or
substance, by solvent extraction, the final step involves the quantitative
determination of the element or substance of interest. Simple colorimetric or,
better, spectrophotometric methods may be applied directly to the solvent
extract utilising the absorption bands of the complex in the ultraviolet or visible
region. A typical example is the determination of nickel as dimethylglyoximate in
ehloroform by measuring the absorption of the complex at 366 nm.

With ion-association complexes, improved results can often be obtained by
developing a chelate complex after extraction. An example is the extraction of
uranyl nitrate from nitric acid into tributyl phosphate and the subsequent
addition of dibenzoylmethane to the solvent to form a soluble coloured chelate.

If direct analysis of the solvent extract is impracticable the element is usually
backwashed into an aqueous phase which can be analysed by standard methods.
Further techniques which may be applied directly to the solvent extract are
flame spectrophotometry and atomic absorption spectrophotometry (AAS). An
example of the former technique is the determination of copper as the
salicylaldoxime complex in chloroform; the organic extract is sprayed directly
into an oxy-acetylene flame and the spectral emission of copper at 324.7 nm is
measured. The direct use of the solvent extract in AAS may be advantageous since
the presence of the organic solvent generally enhances the sensitivity of the
method (Chapter XXII).

Automation of solvent extraction. Although automatic methods of analysis
do not fall within the scope of the present text, it is appropriate to emphasise here
Id! extraction methods offer considerable scope for automation (Refs 2,
. 4). Details of the use of automatic analysers are best obtained by referring to the
manufacturers’ manuals (e.g., Ref. 5) as this is a rapidly expanding field with many
changes occurring.

SOME APPLICATIONS*

BERYLLIUM AS THE ACETYL-
com 1 Discussion. Beryllium forms an acetylacetone

295 n soluble in chloroform, and yields an absorption maximum at

bv ra'^H acetylacetone in the chloroform solution may be removed

thesnl' with O.lM-sodium hydroxide solution. It is advisable to treat

solutinn'tu up to 10 fig of Bc with Up to 10 cm^ of 2 per cent EDTA

• he latter will mask up to 1 mg of Fe, Al, Cr, ,Zn, Cu, Pb, Ag, Ce, and U.

A.R hprTr ^ solution containing 10 ^g of beryllium in 50 cm^ : use

hbeake/ ®^S 04 , 4 H 20 . To 50.0 em^ of this solution contained in

’ ^ ° dilute hydrochloric acid until the pH is 1.0, and then introduce 10.0

rtlier, and more complex, applications will be found in Chapter XVIII.

153

VI, 8 QUANTITATIVE INORGANIC ANALYSIS

cm’ of 2 per cent EDTA solution. Adjust the pH to 7 by the nciciition of 0.1 A/-
sodiiim hydroxide solution. Add 5.0 cm’ of 1 per cent aqueous acclylacetonc and
readjust the pH to 7~8. After standing for 5 minutes, extract the colourless
beryllium complex with three lO-cm’ portions of chloroform. Wash the
chloroform extract rapidly with two 50-cm’ portions of 0. 1 Af-sodium hydroxide
in order to remove the excess of acctylacetonc. To determine the absorbance at
295 nm (in the ultraviolet region of the spectrum) it may be necessary to dilute the.
extract with chloroform. Measure the absorbance using 1.0-cm absorption cells
against a blank.

Repeat the determination with a solution containing 100 ;/g of iron(ni)and of
aluminium ion: the absorbance is unafTcctcd.

VI. 8. DETERMINATION OF BORON USING FERROIN. Discussion.
The method is based upon the complexation of boron as the bis(sali-
cy!ato)boratc(lII) anion (.A). (Borodisalicylatc). and the solvent extraction into
chloroform of the ion-association complex formed with the ferroin.

/ O-.

l !!

\ o

The intensity of the colour of the extract due to ferroin is observed
spectrophotomclrically and may be related by calibration to the boron content of
the sample.

The method has been applied to the determination of boron in river water and
sewage (Ref. 6), the chief sources of interference being coppcrjll) and zinc ions,
and anionic detergents. The latter interfere by forming ion-:issoci:ition complexes
with ferroin whicli arc extracted by chloroform; this property may, however, be
utilised for the joint determination of boron and anionic detergents by the one
procedure. The basis of this joint determination is that the ferroin-anionic
detergent complex may be immediately extracted into chloroform, whereas the
formation of the borodisalicylatc anion from boric acid and .salicylate requires a
reaction time of one hour prior to extraction using ferroin. The absorbance of the
chloroform extract obtained after zero minutes thus gives a measure of the
anionic detergent concentration, whereas the absorbance of the extract after a
one hour reaction period corresponds to the amount of boron plus anionic
detergent present. Interference due to copperfll) ions may be eliminated by
masking with EDTA.

Reagents. Sulphuric acid solution. 0.05Af.

Sodium hydroxide solution, 0.1 A/.

Sodium salicylate solution. 10 per cent w/v.

EDTA solution, I per cent w/v; use the disodium salt of EDTA.

Ferroin solution, 2.5xlO*’AI. Dissolve 0.695 g of iron(ll) sulphate
heptahydratc and 1.485 g of 1,10-phenanthrolinc hydrate in 100 cm’ of distilled
water.

Boric acid solution, 2.5 x 10"* Af. Dissolve 61.8 mg of boric acid in 1 dm’ of
distilled water; dilute 250 cm’ of this solution to 1 dm’ to give the standard boric
acid solution.

154

SOLVENT EXTRACTION VI, 9

Use analytical reagent grade materials whenever possible and store the
solutions in polythene bottles.

. Procedure, (a) Zero minutes reaction time. Neutralise a measured volume
of the sample containing T-2 mg dm"^ of boron with sodium hydroxide or
sulphuric acid (0.0571/) to a pH of 5.5 (use a pH meter). Note the change in volume
and hence calculate the volume correction factor to be applied to the final result.
Measure 100 cm^ of the neutralised sample solution into a flask, add 10 cm^ of 10
percent sodium salicylate solution and 17.5 cm^ of 0.05Ar sulphuric acid, and mix
the solutions thoroughly. Adjust the pH of the solution to pH 6 to 7 with O.IM
sodium hydroxide and transfer the solution immediately to a separating funnel;
wash the flask with 20 cm^ of distilled water and add the washings to the rest of
the solution. Add by pipette 1 cm^ of 1 per cent EDTA solution and 1 cm^ of 2.5
X 10“^M ferroin solution and again throughly mix the solution. Add 50 cm^ of
chloroform and shake the funnel for thirty seconds to mix the phases thoroughly.
Allow the layers to separate and transfer the chloroform layer to another
separating funnel. Wash the chloroform by shaking it vigorously for thirty
seconds with 100 cm^ of water and repeat this process with a second 100 cm^ of
water. Filter the chloroform phase through cotton-wool and measure the
absorbance, Ag, against pure chloroform at 516 nm in a 1 cm cell (this 1 cm cell
reading is used to calculate the boron concentration on the basis of equation (1),
but if the zero minutes reading is to be used for determination of anionic detergent
concentration a 2 cm cell reading is more suitable).

(b) One hour reaction time. Measure a second 100 cm^ of neutralised
sample solution into a flask, add 10 cm^ of 10 per cent sodium salicylate solution
and 17.5 cm^of0.05Msulphuricacidsolution. Mix thesolutions thoroughly, allow
the mixture to stand for one hour and adjust the pH of the solution to 6-7 with
O.IM sodium hydroxide. Now proceed as previously described, under (a), to
obtain the absorbance, A^,. The absorbance. A, to be used in the calculation of
the boron concentration is obtained from the following equation:

(Ag ( 1 )

is determined by repeating procedure (a), i.e., zero minutes, using 100 cm^
° water in place of the sample solution.

a culate the amount of boron present by reference to a calibration graph of
V . ^Sainst boron concentration (mg dm“^). Multiply the result
of thrsam appropriate volume correction factor arising from neutralisation

c"- 25, 50, 75 and 100 cm^ of the standard boric acid

vielH ^ ^ ^ up to 100 cm^ with distilled water; this

soliif' ^ concentration range up to 2.70 mg dm"^. Continue with each
the T under Procedure (b), i.e., one hour reaction time, except that

Constr t of the boron solution to pH 5.5 is not necessary.

concenlSf^ calibration graph of absorbance at 516 nm against boron
carnva For maximum accuracy, the calibration should be

u immediately prior to the analysis of samples.

of copper as the diethyldithio-

reacts with ™ r f^'^cussion. Sodium diethyldithiocarbamate (B)

a slightly acidic or ammoniacal solution of copper(II) in low

155

VI, 10 QUANTITATIVE INORGANIC ANALYSIS

concentration to produce a brown colloidal suspension of the copper(II)
diethyidithiocarbamatc. The suspension may be extracted with an organic
solvent (chloroform, carbon tetrachloride or butyl acetate) and the coloured
extract analysed spcctrophotomctrically at 560 nm (butyl acetate) or 435 nm
(chloroform or carbon tetrachloride).

(C.U,).N-C^ (II)

S-JNn*

Many of the heavy metals give slightly soluble products (some white, some
coloured) with the reagent, most of which arc soluble in the organic solvents
mentioned. The selectivity of the reagent may be improved by the use of masking
agents, particularly EDTA.

The reagent decomposes rapidly in solutions of low pH.

Procedure. Dissolve 0.0393 g of A.R. coppcr(I!) sulphate pcnlahydratc in I
dm-' of water in a graduated flask. Pipette 10.0 cm-' of this solution (containing
about 100 //g Cu) into a beaker, add 5.0 cm-* of 25 per cent aqueous citric acid
.solution, render slightly alkaline with dilute ammonia solution and boil olT the
excess of ammonia; alternatively, adjust to pH 8.5 using a pH meter. Add 15.0
cm’ of 4 per cent EDTA solution and cool to room temperature. Transfer to a
separatory funnel, add 10 cm’ of 0.2 per cent aqueous sodium diethyidithio-
carbamatc solution, and shake for 45 seconds. A yellow-brown colour develops in
the solution. Pipette 20 cm’ of butyl acetate into the funnel and shake for 30
seconds. The organic layer acquires a yellow colour. Cool, shake for 15 seconds
and allow the phases to separate. Remove the lower aqueous layer; add 20 cm’ of
5 per cent sulphuric acid (v/v), shake for 15 seconds, cool, and separate the
organic phase. Determine the absorbance at 560 nm in 1.0-cm absorption cells
against a blank. All the copper is removed in one extraction.

Repeat the experiment in the presence of 1 mg of irontlll); no interference can
be detected.

SOLVENT EXTRACTION VI, 11

VI, 11. DETERMINATION OF IRON BY CHLORIDE EXTRAC-
TION. Discussion. The extraction ofirdn(III) chloride from hydrochloric acid . ‘
with diethyl ether (pro^ly as the solvated complex H[FeCl 4 ]) has long beeiT
known, but the amount of metal extracted depends upon the concentration of the
acid and passes through a maximum at about 6M-hydrochloric acid.

Elements thaf extract well as chloride complexes include Sb(V), As(III), Ga(III),
Ge(lV), T1(III), Hg(ri), Mo(VI), Pt(II), and Au(III). Elements which are partially
extracted include Sb(III), As(V), V(V), Co(II), Sn(II), and Sn(IV). Many solvents
with donor oxygen atoms, including di-isopropyl ether, ^^'-dichlorodiethylether,
ethyl acetate, butyl acetate, and pentyl acetate, have been employed. In most
cases the optimum extraction depends upon the acid concentration.

The extraction of large amounts of iron is conveniently made with iso-butyl
acetate; this solvent has the merit of low volatility and of almost negligible
temperature rise during the extraction (unlike diethyl ether).

To gain experience in the procedure, experimental details are given for the
extraction of iron(III) in hydrochloric acid solution with diethyl ether.

Procedure. Weigh out 16.486 g of A.R. hydrated ammonium iron(III)
sulphate and dissolve it in 250 cm^ of 6M-hydrochloric acid in a graduated flask.
Extract 25.0 cm^ of the iron(III) solution (which contains 200 mg of Fe) with three
25-cm^ portions of pure diethyl ether (1): shake gently for 3 minutes during each
extraction. Combine the three ether extracts and strip the iron from the ether by
shaking with 25 cm^ of water: approximately 99.9 per cent of the iron is removed
by this method. Boil off any ether remaining in the aqueous extract on a water-
bath (caution!), and determine the iron by titration with standard O.IJV-
potassium dichromate after previous reduction to the iron(II) state. The iron
recovered should not be less than 99.6 per cent (2).

Notes. 1. The factors of importance in the diethyl ether extraction of iron
are:

(u) The iron must be in the iron(III) state, since iron(II) chloride is not ex-
tracted.

fl hydrochloric acid concentration must be close to 6M.
w) The extraction should be carried out in subdued light, since ether
Photochemically reduces iron(III).

U The ether should be free from ethanol and peroxides because these reduce
iron(III) chloride.

In ^®iicentration of anions other than chloride should be kept low.

U J Heat is generated by the mixing of the ether and the hydrochloric acid-
iron(III) chloride solution so that cooling of the mixture under the tap or in

ice IS essential.

steel he adapted to the determination of iron in an iron ore or a

weiehpH • details are as follows. Dissolve a 0.5g sample, accurately
by Lat'’ hM-hydrochloric acid and 4 cm^ of concentrated nitric acid

then riie ^ mixture on a water bath. Evaporate the solution to dryness and
to a com° residue in 15 cm^ of 1 : 1 hydrochloric acid. Transfer the solution
Extract tK extractor and rinse the vessel with a little 6M-hydrochloric acid,
until the ®ther or with peroxide-free di-isopropyl ether

solutionofth • ^ solution is colourless. Transfer the ethereal

ethereal sol, r chloride to a separatory funnel, strip the iron from the

Determine °r three washings with an equal volume of water.

ermine the iron content as above.

157

VI, 10 QUANTITATIVE INORGANIC ANALYSIS

concentration to produce a brown colloidal suspension of the copperfll)
diethyidithiocarbamatc. The suspension may be extracted with an organic
solvent (chlorororm, carbon tetrachloride or butyl acetate) and the coloured
extract analysed spcctrophotomctrically at 560 nm (butyl acetate) or 435 nm
(chlorororm or carbon tetrachloride).

( 0 |

S-JNa*

Many of the heavy metals give slightly soluble products (some white, some
coloured) with the reagent, most of which arc soluble in the organic solvents
mentioned. Tlie selectivity of the reagent may be improved by the use of masking
agents, particularly E13TA.

Tlic reagent decomposes rapidly in solutions of low pH.

Procedure. Dissolve 0.0393 gof A.R.copperfllisulphatcpcntahydratein 1
dm-' of water in a graduated flask. Pipette 10.0 cm^ of this solution (containing
about 100 //g Cu) into a beaker, add 5.0 cm-' of 25 per cent aqueous citric add
solution, render slightly alkaline with dilute ammonia solution and boil off the
CXCC.SS of ammonia; altcmativcly, adjust to pH S.5 u.sing a pH meter. Add 15.0
cm-' of 4 per cent EDTA solution and cool to room temperature. Transfer to a
separatory funnel, add 10 cm' of 0.2 per cent aqucou.s sodium diclhyldithio-
carbamatc solution, and shake for 45 seconds. A yellow-brown colour develops in
the solution. Pipette 20 cm-' of butyl acetate into Uie funnel and shake for 30
seconds. The organic layer acquires a yellow colour. Cool, shake for 15 seconds
and allow the phases to separate. Remove the lower aqueous layer; add 20 cm*' of
5 per cent sulphuric acid (v/v). shake for 15 seconds, cool, and separate the
organic phase. Determine the absorbance at 560 nm in 1.0-cm absorption cells
against a blank. All the copper is removed in one extraction.

Repeat the experiment in the presence of 1 mg of ironllll); no interference can
be detected.

A'l, 10. DETERMlN.VriON OF COPPER AS THE ‘NEO-CUPROIN’
COMPLEX. Discussion. ‘_Nco-cuproin' (2.9-dimethyl-l,I0-phenanthrolinc)
can, under certain condition-sTbcKaveinriin almost specific reagent for copperO).
The complex is solublcjai;hlorofbrtTKuitnvbsq!^^ to

thcdctermratitltm of copper inTastTron,''a1fov steels, lead-tin .solder, and various
metals.

Procedure. To lO.Ocm^ of the solution containing up to 2(X)//g of copper in
a separatory funnel, add 5.0 cm^ of 10 per cent hydroxylammonium chloride
solution to reduce Cu(II) to Cu(l), and 10 cm* of a 30 per cent sodium citrate
solution to complex any other metals which may be present. Add ammonia
solution until the pH is about 4 (Congo red paper), followed by 10 cm* of a O.I
per cent solution of “neo-cuproin* in absolute ethanol. Shake for about 30
seconds with 10 cm* of chloroform and allow the layers to separate. Repeat
the extraction with a further 5 cm* of chloroform. Measure the absorbance at
457 nm against a blank on the reagents which have been treated similarly to the
sample.

156

SOLVENT EXTRACTION VI, 13

In general, the ‘primary’ dithizoriates are of greater analytical utility than the ^
‘secondary’ dithizonates, which are less stable and less soluble in organic solvents.

Dithizone is a violet-black solid which is insoluble in water, soluble in dilute
ammonia solution, and also soluble in chloroform and in carbon tetrachloride to
yield green solutions. It is an excellent reagent for the determination of small
(microgram) quantities of many metals, and can be made selective for certain'
metals by resorting to one or more of the following devices :

(a) Adjusting the pH of the solution to be extracted. Thus from dilute acid
solution (0.1-0.5N) silver, mercury, copper, and palladium can be separated from
other metals; bismuth can be extracted from a weakly acidic medium; lead and
zinc from a neutral or faintly alkaline medium; cadmium from a strongly basic
solution containing citrate or tartrate.

(b) Adding a complex-forming agent or masking agent, e.g., cyanide,
thiocyanate, thiosulphate, or EDTA.

It must be emphasised that dithizone is an extremely sensitive reagent and is
applicable to quantities of metals of the order of micrograms. Only the purest
dithizone (e.g., A.R.) may be used, since the reagent tends to oxidise to
diphenylthiocarbadiazone S=C(N=NC 6 H 5 ) 2 : the latter does not react with
metals, is insoluble in ammonia solution, and dissolves in organic solvents to give
yellow or brown solutions. Reagents for use in dithizone methods of analysis
must be of the highest purity (e.g., A.R.). De-ionised water and redistilled acids are
recommended : ammonia solution should be prepared by passing ammonia gas
into water. Weakly basic and neutral solutions can frequently be freed from
reacting heavy metals by extracting them with a fairly strong solution of
dithizone in chloroform until a green extract is obtained. Vessels (of Pyrex)
should be rinsed with dilute acid before use. Blanks must always be run.

Only one example of the use of dithizone in solvent extraction will be given in
order to illustrate the general technique involved.

Procedure. Dissolve 0.0067 g of pure lead chloride in 1 dm^ of water in a
graduated flask. To 10.0 cm^ of this solution (containing about 50 jug of lead)
contained in a 250 cm^ separatory funnel, add 75 cm^ of ammonia-cyanide-
sulphite mixture (1), adjust the pH of the solution to 9.5 (pH meter) by the
cautious addition of hydrochloric acid*, then add 7.5 cm^ of a 0.005 per cent
solution of dithizone in chloroform (2), followed by 17.5 cm^ of chloroform.
Shake for 1 minute, and allow the phases to separate. Determine the absorbance
at 510 nm against a blank solution in a 1.0 cm absorption cell. A further
extraction of the same solution gives zero absorption indicative of the complete
extraction of the lead. Almost the same absorbance is obtained in the presence of
100 Jig of copper ion and 100 jig of zinc ion.

Notes. 1. This solution is prepared by diluting 35 cm^ of concentrated
ammonia solution (sp. gr. 0.88) and 3.0 cm^ of 10 per cent potassium cyanide
solution (caution) to 100 cm^, and then dissolving 0.15 g of sodium sulphite in the
solution.

2. One cm^ of this solution is equivalent to about 20 jug of lead.

It is essential that the pH of the mixture does not fall below 9.5, even temporarily, as there is always

the possibility that HCN could be liberated.

159

VI, 12/13 QUANTITATIVn INORGANIC ANALYSIS

yU 12. DETERMINATION OF IRON AS THE 8-IIYDROXY-
^UINOLATE.* Discussion. Iron(IlI) ( 50-200 /ig) canJlc^xtracted from
HQueous solution with a 1 per rent st^ution of S-hviiimy^uinolinenrchiDr^^
EvTliTnlTlccxIraaion when iTTirpTT o niic aqueolj^lut ionJsJbelH^^
" ^N-Trf-?::c>-v nTcTrc l7cobalt, ccrium(lU). and aluminium do no li^rfcieJ rondfll
oxinatc is fI:irL3^ourcd in BilorolormTancn n^or bs at 4 70 nrn.,

' pfoFMinr . — out 015226 g of A.R. hydrated ammonium ironflll)
sulphate and dissolve it in 1 dm’ of water in a graduated flask; 50 cm’ of this
solution contain 100 ;/g of iron. Place 50.0 cm’ of the solution in a lOO-cm’
separatory funnel, add 10cm’ of a 1 percent oxinc (A.R.) solution in chloroform
and shake for I minute. Separate the chloroform layer. Transfer a portion of the
latter to a l.O-cm absorption cell. Determine the absorbance at 470 nm in a
spectrophotometer, using the solvent as a blank or reference. Repeat the
extraction with a further 10 cm’ of I per cent oxinc solution in chloroform, and
measure the absorbance to confirm that all the iron was extracted.

Repeat the experiment using 50.0 cm’ of the ironflll). solution in the presence of
100 //g of aluminium ion and 100 //g of nickel ion at pH 2.0 (use a pH meter to
adjust the acidity) and measure the absorbance. Confirm that an clTcctivc
separation has been achieved.

Note, Some typical results are given below. Absorbance after first
extraction 0.605; after second extraction 0.004; in presence of 100 fig A1 and 100
fig Ni the absorbance obtained is 0.602.

VI, 13. DETERNDNATION OF LEAD BY THE DITHIZONE
METHOD.t Discussiott. Diphcnylthiocarbazonc (dithizonc) behaves in
solution as a tautomeric mixture of(C)and (D):

N-NHC^Il, NII-NIIC.Hj

rt /

HS-C S=C

\ \

N=NCdIs N^NCJI,

(Q (ID

It functions as a monoprolic acid (pA'„ = 4.7) up to a pH of about 12; the acid
proton is that of the thiol group in (C). ‘Primary’ metal dithizonates are formed
according to the reaction .

+nH3Dz5±M(HD7.)„4-nir

Some metals, notably copper, silver, gold, mercury, bismuth, and palladium,
form a second complex (which we may term 'secondary' dithizonates) at a higher
pH range or with a deficiency of the reagent ;

2M (H Dz)„ 5^ M , Dz„ + n H j Dz

In tilts and the .succeeding determinations of minute quamitieu of the various elements involving
sohcntextractionandspeclrophotomctric analysis of the solvent extracts, some simple experiments
with known microgram quantities of tlic elements will be described. These will illustrate the
principles involved; they can be readily ndapicd to the delermimition of unknown solutions by
utilising a calibration graph (absorbance-concentration in /;g).

T IS experiment is not recommended for elementary students or students having little experience of
analytical work.

158

SOLVENT EXTRACTION VI, 13

In general, the ‘priiriary’ dithizonates are of greater analytical utility than the
‘secondary’ dithizonates, which are less stable and less soluble in organic solvents. '

Dithizone is a violet-black solid which is insoluble in water, soluble in dilute
ammonia solution, and also soluble in chloroform and in carbon tetrachloride to
yield green solutions.. It is an excellent reagent for the determination of small
(microgram) quantities of many metals, and can be made selective for certain"
nietals by resorting to one or more of the following devices ;

(fl) Adjusting the pH of the solution to be extracted. Thus from dilute acid
solution (0.1-0.5iV) silver, mercury, copper, and palladium can be separated from
other metals; bismuth can be extracted from a weakly acidic medium; lead and
zinc from a neutral or faintly alkaline medium; cadmium from a strongly basic
solution containing citrate or tartrate.

(b) Adding a complex-forming agent or masking agent, e.g., cyanide,
thiocyanate, thiosulphate, or EDT A.

It must be emphasised that dithizone is an extremely sensitive reagent and is
applicable to quantities of metals of the order of micrograms. Only the purest
dithizone (e.g., A.R.) may be used, since the reagent tends to oxidise to
diphenylthiocarbadiazone S=C(N=NC 6 H 5 ) 2 : the latter does not react with
metals, is insoluble in ammonia solution, and dissolves in organic solvents to give
yellow or brown solutions. Reagents for use in dithizone methods of analysis
must be of the highest purity (e.g., A.R.). De-ionised water and redistilled acids are
recommended: ammonia solution should be prepared by passing ammonia gas
into water. Weakly basic and neutral solutions can frequently be freed from
reacting heavy metals by extracting them with a fairly strong solution of
dithizone in chloroform until a green extract is obtained. Vessels (of Pyrex)
should be rinsed with dilute acid before use. Blanks must always be run.

Only one example of the use of dithizone in solvent extraction will be given in
order to illustrate the general technique involved.

Procedure. Dissolve 0.0067 g of pure lead chloride in 1 dm^ of water in a
graduated flask. To 10.0 cm^ of this solution (containing about 50 /ug of lead)
contained in a 250 cm^ separatory funnel, add 75 cm^ of ammonia-cyanide-
sulphite mixture (1), adjust the pH of the solution to 9.5 (pH meter) by the
cautious addition of hydrochloric acid*, then add 7.5 cm^ of a 0.005 per cent
dithizone in chloroform (2), followed by 17.5 cm^ of chloroform,
hake for 1 minute, and allow the phases to separate. Determine the absorbance
3 10 nm against a blank solution in a 1.0 cm absorption cell. A further
extraction of the same solution gives zero absorption indicative of the complete
extraction of the lead. Almost the same absorbance is obtained in the presence of
iW/ig of copper ion and 100 /rg of zinc ion.

am h This solution is prepared by diluting 35 cm^ of concentrated

i^onia solution (sp. gr. 0.88) and 3.0 cm^ of 10 per cent potassium cyanide

solution ^nd then dissolving 0. 1 5 g of sodium sulphite in the

One cm ofthis solution is equivalent to about 20 /rg of lead.

mixture does not fall below 9.5, even temporarily, as there is always
V ssioaity that HCN could be liberated.

159

VI, 14, QUANTITATIVE INORGANIC ANALYSIS

I, 14. DETERMINATION OF MOLYBDENUM BY THE THIOCY-
ANATE METHOD. Disctission. MolvbdcntimfVD in acid solution when
with tinfll) chloride (best in the p^scfice of a little iron(ll) ionTlT
rnnvi-rtcd In rpely into molvb dcnumivirt his forms a complex with thiocyanate
JonTprobably largel y Mo(SCNLiwHic]i is red in colour. The latter may be
extracted with solvents posscssin^onor oxyg en atom s (3-mcthi? butanoi is
preferred). The colour depends upon “the acid concentration (opiirntnn
concentration 1 A/) and tlieconccnirationofthc thiocyanate ion 1 percent, but
colour inten-sity is constant in the range 2-10 per cent); it is little influenced by
CXCC.S.S of tinfll) chloride. The molybdenum complex has maximum absorption at
465 nm.

Reagent-S. Staitdard molybdenum sphiiion. Di.ssolvc 0.184 g of A.R.
ammonium molylxiate (NH4)f,(Mo7024l4H20 in 1 dm^ of di.stillcd water in a
graduated flask; this gives a 0.001 percent Mo solution containing 10/igMopcr
cm\ Alternatively, dis.solvc 0.150 g of A.R. molybdenum trioxide in a fewem’^ of
dilute .sodium hydroxide solution, dilute with water to about 100 dm^ render
slightly acidic with dilute hydrochloric acid, and then dilute to 1 dny* with water
in a graduated flask: this is a 0,0100 percent solution. It can be diluted to 0.001
per cent wath 0. 1 A/-hydrochloric acid.

Ammonium Iron sulphate solution. Dis.soIvc lOg of the A.R. salt in 100
cm^ of very dilute sulphuric acid.

71n{//) chloride saltitioii. Dissolve 10 g of A.R. tin(II) chloride dihydnitc in
lOOcm^ of hVf-hydrochloric acid.

Potassium thiocyanate solution. Prepare a 10 per cent aqueous .solution
from the A.R. salt.

Procedure. Construct a cjilibration curve by placing 1.0, 2.0. 3.0. 4.0, and
5.0 cm^ of the 0.001 percent Mo solution (containing lO/^g, 20//g. 30//g, 40/ig,
and 50 pg Mo) severally in 50-cm’ separatory funnels and diluting each with an
equal volume of water. Add to each funnel 2.0 cm* of concentrated hydrochloric
acid, 1.0 cm* of the ammonium iron(II) sulphate solution, and 3.0 cm* of the
potassium thiocyanate solution; shake gently and then introduce 3.0 cm* of the
tin(ll) chloride solution. Add water to bring the total volume in each separatory
funnel to 25 cm* and mix. Pipette 10.0 cm* of redistilled 3-mcthylbutanol into
each funnel and shake individually for 30 seconds. Allow the pha,scs to separate,
and carefully run out the lower aqueous layer. Remove the glass stopper and pour
the alcoholic extract through a small plug of purified glass wool in a small funnel
and collect the organic extract in a 1.0-cm absorption cell. Measure the
absorbance at 465 nm in a spectrophotometer against a 3-methylbutanol blank.
Plot absorbance against pg of Mo. A straight line is obtained over the range 0-50
pg Mo: Beer’s law is obeyed.

Determine the concentration of Mo in unknown samples supplied and
containing less than 50 pg Mo per 10 cm*: use the calibration curve, and subject
the unknown to the same treatment as the standard solutions.

The above procedure may be adapted to the determination of molybdenum in
.steel. Dissolve a 1.00-g sample of the steel (accurately weighed) in 5 cm* of 1:1
hydrochloric acid and 15 cm* of 70 per cent perchloric acid. Heat the solution
until dense fumes are evolved and then for 6-7 minutes longer. Cool, add 20 cm*
of water, and warm to dissolve all salts. Dilute the resulting cooled solution to

160

SOLVENT EXTRACTION VI, 15

volume in a 1-dm^ flask. Pipette 10.0 cm^ of the diluted solution into a 50-cm
separatory funnel, add 3 cm^ of the tin(II) chloride solution, and continue as
detailed above. Measure the absorbance of the extract at 465 nm with a
spectrophotometer, and compare this value with that obtained with known
amounts of molybdenum. Use the calibration curve prepared with equal
amounts of iron and varying quantities of molybdenum. If preferred, a mixture of
3 -methylbutanol and carbon tetrachloride, which is heavier than water, can be

used as extractant. . , . ,

Note. Under the above conditions of determination the following elements
interfere in the amount specified when the amount of Mo is 10 //g (error greater
than 3 per cent): V, 0.4 mg, yellow colour (interference prevented by washing
extract with tin(II) chloride solution); Cr(VI), 2 mg, purple colour; W(YI), 0.15
mg, yellow colour; Co, 12 mg, slight green colour; Cu, 5 mg; Pb, 10 mg; Ti(III), 30
mg (in presence of sodium fluoride).

yif 15. DETERMINATION OF NICKEL AS THE DIMETHYLGLY-
-"OXIME COMPLEX. Discussion. Nickel (200-400 nel forms the red dimethyl-
glyoxime complex in a slightly alkaline medium: it is only slightly soluble ^n
chloroform (35-50 ng.Ni cm~^ l. The optimum pH range for extraction of tfie^
n ickel complex is 7-12 in the presence of citrate. The nickel complex absorbs a T
366 nm and also at 465- 470 nm. ~ ~ ~

Procedure. Weigh out 0.135 g of pure ammonium nickel sulphate
(NiS 04 ,(NH 4 ) 2 S 04 , 6 H 20 ) and dissolve it in 1 dm^ ofwater in a graduated flask.,
Transfer 10.0 cm^ of this solution (Ni content about 200 /ig) to a beaker
containing 90 cm^ of water, add 5.0 g of A!R. citric acid, and then dilute ammonia
solution until the pH is 7.5. Cool and transfer to a separatory funnel, add 20 cm^
of dimethylglyoxime solution ( 1 ) and, after standing for a minute or two, 12 cm^
of chloroform. Shake for 1 minute, allow the phases to settle out, separate the red
chloroform layer, and determine the absorbance at 366 nm in a 1.0-cm absorption
cell against a blank. Extract with a further 12 cm^ of chloroform and measure the
absorbance of the extract at 366 nm; very little nickel will be found.

Repeat the experiment in the presence of 500 fig of iron(III) and 500 /ig of
aluminium ion; no interference will be detected.

Note. 1 . The dimethylglyoxime reagent is prepared by dissolving 0.50 g of
A.R. dimethylglyoxime in 250 cm^ of ammonia solution and diluting to 500 cm^
with water.

Note. 2. Cobalt forms a brown soluble dimethylglyoxime complex which
is very slightly extracted by chloroform ; the amount is only significant if large
amounts of Co (>2-3 mg) are present. If Co is suspected it is best to wash the
organic extract with ca. 0.5M-ammonia solution: enough reagent must be added
to react with the Co and leave an excess for the Ni. Large amounts of cobalt may
be removed by oxidising with hydrogen peroxide, complexing with ammonium
thiocyanate (as a 60 per cent aqueous solution), and extracting the compound
with a pentyl alcohol-diethylether (3:1) mixture. Copper(II) is extracted to a
small extent, and is removed from the extract by shaking with 0.5M-ammonia
solution. Copper in considerable amounts is not extracted if it is complexed with
thiosulphate at pH 6.5. Much Mn tends to inhibit the extraction of Ni; this
difficulty is overcome by the addition of hydroxylammonium chloride. Iron(III)
does not interfere.

VI, 16 QUANTITATIVE INORGANIC ANALYSIS

VI, 16. DETERMINATION OF SILVER BY EXTRACTION AS ITS ION-
ASSOCIATION COMPLEX WITH 1,10-PHENANTHROLINE AND
BROMOPYROGALLOL RED. Discussion. Silver can be extracted from a
nearly neutral aqueous solution into nitrobenzene as the blue ternary ion-
association complex fonned between silverfi) ions, 1,10-phenantliroline and
bromopyrogallol red. The method is highly selective in the presence of EDTA,
bromide' and mcrcury(II) ions as masking agents and only thiosulphate appears
to interfere {Ref. 7).

Reagents. Silver nitrate solution. 10“'*Af. Prepare by dilution of a
standard 0.1 A/ silver nitrate solution.

l,JO-Phcnanthrolinc .solution. Dissolve 49.60 mg of analytical grade 1,10-
phcnanthrolinc in distilled water and dilute to 250 cm-\

Ammonium acetate solution, 20 per cent. Dissolve 20 g of the analytical grade
salt in distilled water and dilute to 100 cm^.

Bromopyrogallol red solution. 10“‘‘Af. Dissolve 14.0 mg of bromopyrogallol
red and 2.5 g of ammonium acetate in distilled water and dilute to 250 cm\This
solution should be discarded after five day.s.

EDTA .solution, 10* 'Af. Dissolve 3.7225 g of analytical grade disodium salt in
distilled water and dilute to 100 cm'*.

Sodium nitrate solution, 1 A/. Dissolve S.5 g of analytical grade sodium nitrate
in distilled water and dilute to 100 cm^.

Nitrobenzene, analytical grade.

Sodium hydroxide, analytical grade pellets.

Procedure, (n) Calibration. Pipette successively 1, 2, 3, 4 and 5 cm^ of
10 A/ silver nitrate solution, I cm^ of 20 per cent ammonium acetate solution, 5
cm^ of lO'^Af 1,10-phcnanthrolinc solution. 1 cm^ of 10* 'A/ EDTA solution
and I cm^ of l.M sodium nitrate solution into five 100 cm-' separating funnels.
Add sufiicient distilled water to gis'c the same volume of solution in each funnel,
then add 20 cm^ of nitrobenzene and shake by continuous inversion for one
minute. Allow about ten minutes for the layers to separate, then transfer the lower
organic layers to different 1 00 ern ' separating funnels and add to the latter 25 cm^
of 10* ■‘A/ bromopyrogallol red .solution. Again .shake by continuous inversion
for one minute and allow about thirty minutes for the layers to separate. Run the
lower nitrobenzene layers into clean, dry 100 cm’ beakers and swirl each beaker
until all cloudiness disappears (Note 1). Finally transfer the solutions to 1 cm cells
and measure the absorbance at S^'-O nm against a blank carried through the same
procedure but containing no ' Ivcr. Plot a calibration curve of absorbance
against silver content (//g).

1 cm’ of lO-'Af AgNO., = 10.788 //gof Ag

(h) Determination. To an aliquot of the silverf!) solution containing
between 10 and 50 gg of silver, add sufficient EDTA to complex all those cations
present which form an EDTA complex. If gold is present (:j- 250 pg) it is masked
by adding sufiicient bromide ion to form the AuBr 4 ~ complex. Cyanide,
thiocyanate or iodide ions arc masked by adding sufiicient mercury(II) ions to
complex these anions followed by sufficient EDTA to complex any c,xccss
mcrcury(ll). Add I cm’ of 20 per cent ammonium acetate solution, etc., and
proceed as described under Calibration.

Note. I . More rapid clarification of the nitrobenzene extract is obtained if
the beakers contain about 5 pellets of sodium hydroxide. The latter is, however, a

162

SOLVENT EXTRACTION VI, 17/18/19

source of instability of the- colour system and its use is therefore not
recomi^nded. ’

VI, /n. DETERMINATION OF URANIUM AS THE 8-HYDROXY-
bwNOLATE. Discussion. UrahiumfVIl may be determined as the 8-
hydroxyguinnlate in concent rations up to 900 /ig at a pH of 8.8 in the presence of
a little EPTA; the yellow oxinate complex absorbs at 400 nm. , Many interfering
~ elements (e.g., iron and aluminium but not titanium) may be masked~~5y

’ increasing the quaiitilv of FOTA in iht-, sulnhuii. "

Simple experimental details follow: these are designed to giye the student
experience in the method. Absorbances are read against a blank solution using
1.0 cm absorption cells.

Procedure. Weigh out 0.106 g of A.R. uranyl nitrate U 02 (N 03 ) 2 , 6 H 20
and dissolye it in 1 dm^ of water in a graduated flask. Mix 10.0 cm^ of the solution
(containing about 500 ^g ofU) with 5.0 cm^ of 0.02M-EDTA, adjust the pH to 8.8
and dilute to 100 cm^ Extract with two 10-cm^ portions of 1.0 per cent oxine
(A.R.) solution in chloroform, and measure the absorbance after each extraction.

Note. Some typical results with a spectrophotometer were ; absorbance at
first extraction, 0.510; at second extraction, 0.005.

Repeat the experiment in the presence of 500 //g of iron (as A.R. iron(III) alum)
and 500 fig of aluminium (as A.R. ammonium alum): it will be found that the
absorbance has increased, suggesting interference from these elements. Increase
the volume of 0.02M-EDTA solution by 5-cm^ portions until the absorbance is
identical with that of the original uranium solution; about 20 cm^ will be
required.

VI, 18. References

1 . H. Irving and R. J. P. Williams (1949). ‘Metal Complexes and Partition Equilibria’, J.
Chem. Soc., 1841.

2. J. Dunbar (1963), in Proceedings ofTechnicon Symposium on Automated Analytical
Chemistry, 130.

3. J. M. Carter and G. Nickless (1970). ‘A Solvent-extraction Technique with the
Technicon AutoAnalyser’, Analyst, 95, 148.

4. F. Trowell (1969). ‘Automated Solvent Extraction’, Laboratory Practice, 18, 44.

5. ‘Automating Manual Methods using Technicon AutoAnalyser 11 System Tech-
niques’, Manual TN1-0I70-01. New York; Technicon Instruments Corporation,
1972.

6. J. Bassett and P. J. Matthews (1974). ‘A Spectrophotometric Method for the
Determination of Boron in Water by the use of Ferroin’, Analyst, 99, 1.

7. R. M. Dagnall and T. A. West (1964). ‘A Selective Extraction System for Trace
Amounts of Silver’, Talanta, 11, 1627.

8. A. K. Babko (1968). Talanta, 15, 721.

VI, 19. Selected bibliography

1. F. D. Snell and C. T. Snell. Colorimetric Methods of Analysis. Vol. II. Inorganic

(1949). Vol.IIA (1959). New York; Van Nostrand. inorganic

2. H. Irving (1951). ‘Solvent Extraction and its Applications to Inorganic Analysis’

Quarterly Reviews, S,2QQ. ’

3. H. Freiser (1952). ‘The Stability of Metal Chelates in Relation to their Use in
Analysis , Analyst, 77, 830.

163

VI, 19 QUANTITATIVE INORGANIC ANALYSIS

4, G. Chariot and D. Bezier (translated by R. C. Murray) (1957). Qtimuiiativc [nors;anic
Analysis. Chapter XIII. Reactions in the Presence of Two Immiscible Solvents.
Separation by Extraction. London : Methuen.

5. G. H. Morrison and H. Preiser (1957). Solvent Extraeiion in Analytical Chemistry.
New York ; John Wiley.

-> 6. E. B. SandcII (1959). Cnloriniciric Determination of Traces of Metals. 3rd edn. New
York: Intcrsciencc Publrshers.

7. L. Alders (1959). l.iquid- Liquid Extraction, Theory and luihoratory Practice. 2nd edn.
Amsterdam: Elsevier Piiblishinc Co.

8. L. C. Craig (195fi). ‘Extraction’, .-inalytical Chemistry. 28, 72.7,

9. G, H. Morrison and H. Preiser. ’Extraction’. .Analytical Chemistry, 1958, 30, 633:
1960.32,37R.

10. .A. K. De, S, M. Khopkar and R. A. Chalmers (1970). Solvent E.xtraction of Metals.
London: Van Nostrand Reinhold Co.

11. J. Slary (1964). The Solvent Extraeiion of .Metal Chelates. Oxford ; Pergamon Press.

12. Y. Marcus (ed.) (1971). Sohent Extraction Reviews. Vol. 1. New York: Marcel
Dekker Inc. Ltd.

13. H. Irving and R. J. P. Williams (1961), ’Liquid-Liquid Extraction’, in Treatise on
Analytical Chemistry, (ed. 1. KollliolT and P. Elvinp) P.irt 1. Vol. 3. New York:
Intcrsciencc.

14. R. M. Diamond and D. G. Tuck (1960). ’Extraction of Inorganic Compounds into
Organic Solvents’, in P. A. Cotton (ed.). Progress in Inoryanie Chemistry Vol. 2. New
York; Intcrsciencc.

1 5. Yu. A. Zolotov (1970). (trans. J. Schmorak). Extraction of Chelate Compounds. Ann
.Arbor; Ann Arbor Science Publishers Inc.

16. A. S. Kertes ;ind ■)'. Marcus (ed.) (1970). Solvent Extraction Rr.search. Proceedings of
the Fifth International Conference on Solvent Extraction Chemistry. Chichester: John
Wiley and Sons. Sussex.

164

CHAPTER VII lON EXCHANGE

Vn, 1. GENERAL DISCUSSION. The term ion exchange is generally
understood to mean th e exchange of ions of like sign between a solution and^
solid highly insoluble body in contact with it. T he solid (ion exchanger) mustTof
course, contain ions of its own, and tor the exchange to proceed sufficiently,
rapidly and extensively to be of practical value, the solid must have an open,
permeable molecular structure so that ions and solvent molecules can'move freel^T
in and out. Many substances, both natural (e.g., certain clay minerals) and
' artihcial, have ion exchanging properties, but for analytical work synthetic -
organic ion exchangers are chiefly of interest, although some inorganic materials,
e.g., zirconyl phosphate and ammonium 12-molybdophosphate, also possess
useful ion exchange capabilities and have specialised applications (Ref. 1 ). All ion
exchangers o f value in analysis have several properties in common: they are •
almost insoluble in water and in organic solvents, and they contain active or

counter ions that will exchange reversibly with other ions in a surrounding

solution witnout any appreciable physical change occurring in the material. The .
~T6n excnanger is oi complex nature and is. m tac t, polymeric. T he polymer carries
"ah electric cnarge that is exactly neutralised by tfifc charges on the counter ions.
These active ions are cations in a cation exchanger and anions in an anion
exchangerTT hus a cation exchanger consists of a polymeric anion and active
jiations, while an anion exchanger is a polymeric cation with active anions. ' ■ -

A widely use n cation excnange resin is that obtained by th e copolymerisalion

pfr styrene | CH=CH 2 | and a small proportion of divinylbenzene

followed by sulphonation; it may be

CH,=CH—

CH=CH2 ,

represented as

Vn, 1 OUANTITATiVE INORGANIC ANALYSIS

Tlie formula enables us to visualise a typical cation exchange resin. It consists of a
polymeric skeleton, held together by linkings crossing from one polymer chain to
the next: the ion exchange groups arc carried on this skeleton. The physical
properties arc largely determined by the degree of cross-linking. This cannot be
determined directly in the resin ilscif: it is often specified as the mole per cent of
the cross-linking agent in the mixture polymerised. Thus ‘polystyrene sulphonic
acid. 5 per cent DVB' refers to a resin containing nominally 1 mole in 20 of
divinylbctv/cne; the true degree of cross-linking probably differs somewhat from
the nimiinal value, but the latter is nevertheless useful for grading resins . Highly
cross-linkcti resins arc generally more brittle, harder, and more impervious than .
the lightly cross-linked materials; the prelercnce oFTi resin for one ion over
another is inllueticed by the' degree of cross linking. The solid g ranul es of resin
swell wh en placed in water, but the swelling Js linrn^d bjjjic.carsilGiTting. In the
bove'exampie the tlivTnvlbenyene units ‘wclB’ the nolvstyrene cht iinsJLQgelhci:
and iirevenl it from swelling mdeliniteiy and dispersing. The rc.su lting structure is
a va.st snon cc-like ne luor>:_w itli neg atively charged suljihcmj!Tcji£n£Jdia cited
fi rmly to the framework. Tl ijksx;_Ji\_t^jnc£aj2y£]5> a rg es arc ha Ian ced by a n

eumvalent n umber of cations; hydrogen ioirsjn^hc.IixtlcogCfiJo.rmjpfTl^
and sodium tons hrTh?''sOdTiTTT rfonn~un'fic^ resin, etc. These ions move ffeely*
Wiilim the wate r-lillcU po res and arc sometimes called mobile ions; Tlw are tluT '
ions w inch c.xcTi.llti’c wlllftnlier ioits/pVhen a cation exchanger con fa in i n^" mobile
umsC^ IS brought mioTontact wiiit a solution containing cmiions A ’ the latter
dilTuse into the resin structure and cations C* diffuse out until equilibrium is
attained. The solid and the solution then contain both cations C‘ and A" in
numbers depending upon the position of equilibrium. TItc same mechanism
operates for the exchange of anions in an anion exchanger.

Anion exchangers are likcvusc cross-linked. high-im>1ecular-wcight p olymers.
Tlieirbasj£_chaiWieEjLtlu c, to the presence of ammo, -subsii ui a’d-amillfCSr
jiuatcrnarv ammonium nrotips. The polymers containing quaternary ammonium
g ronps arc stronc b ases: liios c wnti amino or substitute d amino g roups possess
weak basic properties. A w ideh used amon e\ch;in£^c^csixL i.s~m~cpafed by
copolyineXi^i UiQn of styrene and a little divinylbcn/.cnc, followed b y chloro-.
mciInlnlioiL fintroduction of the — CH.Cl grouping, say, in the free para'
iiid interactio n with a base such as trimethyiaminc. A hypothetical

~TuidTionT

formulation of siich a polystyrene anion exchange resin is gii en on p. 167.

Nuineroti.s types of both cation and anion exchange resins itavc been prepared,
but only a few can be mentioned here Cation e xchang e resins includ e Uiat
JJI^Pared by the copolymcrisation of methacrvlic acid Cit,~ClCIlj)— .CQOTl
“ ^^'tkrglyxoTTismetha

CH;=C(C!Ij)-COOril,

CH,=C(CM,,)— COOCH.

.(as the cr oss-linkine acent ): this contains free — COOH groups and has weak
acidic properties. Weak cationTltchange.rc.sins containing free -r-COOH and
~“9i!_8t^ups have also been synthesi.scd. Anion exchange resins containing
primary; secondary, or tertiary amino groups possc.ssXveakly basic properties.

We may define a cation exchange rcsinais a high molecular \veiaht, CLOS&:.linkccL
4IQbLmer_contajnin^ujphop ct ercrou^as an integ ral

.naryofthe j^esin and an equivalent amouqt of cution^ ^fan anion exchange resin is a

166

ION EXCHANGE VH, 2

and an equivalent amount of anions such as chloride, hydrox yl, or sulphate ions.
The fundamental requirements of a useful resin a re: ~ ~~

le resin must h e su'tncienti v cross-hnked to have only a neghaMe SDlxtbility .
le resin must be sufficiently hydrophilic to permit diffusion of ions through
le structure at a finite and usable rate.

'57 The resin must contain a sufficient number of accessib le ionic exchange groups
must be chemically stable . '

— 4r'TlTe swollen resin must be denser than water .

Some of the commercially available ion exchange resins are collected in Table
VII, 1. These resins, produced by different manufacturers, are: often inter-
changeable, and similar types will generally behave in a similar manner. The
reader will find the table useful if it is desired to repeat original work carried out
with a resin which, for some reason, is not immediately available.

VII, 2. . ACTION OF ION EXCHANGE RESINS- Cation exchange resins*
contain free cations which can be exchanged for cat ions in solution (soln).

( Res.A~)B'*' +0'*^ (solution) ::^(Res.A~ 1C'*' +B'*'(soln) , < (1)

If the experimental conditions are such that the equilibrium is completely
displaced from left to right the ion C*" is completely fixed on the cation exchanger.
If the solution contains several ions (C"^, D'*', and E'*') the exchanger may show
different affinities for them, thus making separations possible. A typica l example
i s the displacement of sodium ions in a sulphonate resin by calcium ion^

2 (Res.S 03 “)Na++Ca 2 +(soln) :;±(Res.S 03 )“ 2 Ca^"' +2Na+(soln) (2)

The reactio n is reversible :-b v passing a solution containing sodium ions through
the product, the calcium ions may be removed from the resin and the original
sodium form regenerated. Similarlyy/by passing a solution of a neutral salt

• These will be represented by Qt6S; A~)B:*^, where Res. is the basic polymer of the resin. A’ is the
anion attached to the polymeric framework..Blis abe active or mob ile cation: thus a sulphonated
polystyrene resin in the hydrogen form would be writtenas(Res.SOnH+: A similar nomenclature
will be employed for anion exchange resins, e.g., (Res.NMcj'^lCl'.

Table VH. I. Comparable ion exchange materials

•s

i c;

Si E

i £

C ^

la'

»-

«s

— fs

C3 ■<

ir.

O OO

I 3 =

I’g

o o

O O
(A tA

C C

r*-i, V.

u o

E ;-i

a u

i;i;z

5 6 H

^ %n

CiU » DCC OC

c o

r t cr,
c: c:

V V
X; ^

H 5

*? -

*< ■<
U L»

3 C3

C Q

o o

5 H

o t?

U c
—

.■?: c .

"3:1

vTi V'.

— rl
ri r-i

n X US x:

, V V

\ N NJ

n: n

VJ O t>

£ x:

r; *>'

-»> c:
Vi fS

168

resms

ION EXCHANGE VH, 2

through the hydrogen form of a sulphonic resin, an equivalent quantity of the
corresponding acid is produced by the following typical reaction :

(Res.S03")H+ +Na+Cr (soln) ^(Res.S03~)Na+ +H+a- (soln)

For the strongly acidic cation, exchange resins; such as the cross-linked
'polystyrene sulphonTc acid resins, the exchange capacity is virtually independenT
of the pH of the solution. F or weak acid cation exchangers, such as those
' containing the carboxylate group, ionisation o ccurs to an appreciable exte nt only
m alk'aline solution, i.e ., in their salt foniTTconsequently the carboxylic resins nave ~
' Wy little action in solutions below pH 7 . T hese carboxylic exchangers in the
hydrogen form will absorb strong oases Irotn solution:

(Res.COO~)H+ +Na+OH' (soln)^(Res.COO-)Na+.+ H20

( 4 )

but will have little action upon, say, sodium chloride; hydrolysis of the salt form -
of the resin occurs so that the base may not be completely absorbed even if an
excess of resin is present.

Strongly basic anion exchange resins, e.g., a cross-linked poly styrene
containing quaternary ammonium groups, are largely ionised in both t he
"hydroxide and the salt forms. Some of their typical reactions may be represented
as:

2(Res.NMe3+)Cr -1-8042 " (soIn);?i(Res.NMe3+)2S042- -|-2Cr (soln)

( 5 )

(Res.NMe3+)Cl" -f OH" (soln)^(Res.NMe3+)OH- -f Cl" (soln)

( 6 )

(Res.NMe3+)OH- -f H+Cl" (soln)^(Res.NMe3+)Cr -fH 20

( 7 )

These resins are similar to the sulphonate cation exchange resins in their actiyity,
and their a ction is largely independent of pH. Weakly basic ion exchange resins
contain little of the hyd roxide form in basic solution. 1 he equilibrium of, say;

(Res.NMe2)-fH20^(Res.NHMe2)+0H- (8)

is mainly to the left and the resin is largely in the amine form. This may also be
expressed by stating thaOn basic solution the free base Res.biHMe2-OH is very
little ionised. In acidic solution, however, they behave like the strongly basic ion
exchangeresins,yieldingthehighlyionisedsaltform;

(Res.NMe 2 )-(-H+Cr ^(Res.NHMe 2 '^)Cr • (9)

T hey can be ., used -i a . a s.id solution for the exchange of anions, fo r example :
(Res.NHMe2+)Cl--f NO3- (soln) :?i(Res.NHMe2+)N03- 4- Cr (soln)

( 10 )

Basic resins in the salt form are readily regenerated with alkali...

~ Ion exchange- equilibriaT^ The ion- exchange proces*s, inyolying the
replacement of the exchangeable ions in the resin by ions of like charge from
a solution, may be written:

^R + Bs^Bf^ + As

169

VII, 2 QUANTITATIVE INORGANIC ANALYSIS

Tlic process is a reversible one. The extent to. which one ion is absorbed in
preference to another is oi tuirdamcntal importance, it will determine the
readiness with which two or more substances, whicli form ions of like charge, can
be separated by ion exchange and also the case with which the ions can
subsequently be removed from the resin. The fac ^s dcterminingi hcjjjsllibtilign
of ions between an ion excli ancc resin aiuTas oIution include^

low aqucoiis'^ncentrations and at

ordinary lenTpcratiires th e cxi^ofcxchnnncip.crea!ies with increasi^valcncy of '
the exchanging ion , i.e., ^

Na-^ <Ca-^ < AP" <Th‘‘ ^

(h) Under similar conditions and/cfinsiant valence, for jinivalcnt ions the
£xtcnt of cxchan.gc increases with idccrcasc in s iy c of the'TiydriItcg~canon L i^ <
Tr”< Na < Nlij K”’ < Cs ’ , whil e fi^ivalent ionTihe ion ic'sixe is

an imporoint facinrbiii ihcincomplctcdi.ssociation ofsalts ofbivalenl mefalsTifa)

Cd-‘ < Be’" < Mn’" < Mg=" = Zn* * < Cu’*

= Nr" < Co'" < Ca*" < Sr’" < Pb^‘ < Ba^".

(c) With strongly basic anion exchange resins, the extent of c.xchangc for
univalent anions saries with the sire of hydrated ion in a .similar manner to that
indicated for cations. In dilute solution polyvalent anions arc generally absorbed
preferentially.

id) When a cation in solution is being e.xchangcd foran ion ofdifTerent valency
the relative alTinity of the higher valent ion increases in direct proportion to the
dilution. Thus to exchange a higher valent ion on the exchanger for one of lower
valency in solution, exchange will be favoured by increasing the c*oncentration,
while if the lower valent ion is in the exchanger and the higlicr valent ion is in
solution, exchange will be favoured by high dilutions.

(ii) Ndfurc ofioti t'-vc/uiiigc rc-siii. The absorption* of ions will depend upon the
nature of the functional groups in the resin. It will also depend upon the degree of
cross-link ing ; as the deerce of c j.Q£S-lin|;i rtg is incrensed. resins become more .
selective towards io ns of different sir.e^the volume ofThc ion is assumed to
include the waierlir li\xIraTioh),rt4HCimxavilhJ]ie-MiuiI]ixAydiataLy plunic will
u|willy be absorbed preferentially. *

foiTexeliange capacity. Ttrc total ion exchange capacity of a resin is
ji£E£ndcnJL.upj3ii-ilic-.iQtal_Jiuriil3ec,<iirion-activc groups per uiiiTTvetghT' of

_ material, an d the greater the number of ions, the preaV^Tvill birtlie^capacTty yTlie”
total ion exclHmiiC caponly is usual ly expressed as milli-eqiiiv.alents per cram of
exchanger: i t may be regarded as an equivalent weight, tlie latter being the
reciprocal oT the former, i,c., mcq. per g - lOOO/oquivalcnt wciclit . The capacities
.ofthc _wcaklv acidic nn iLweaklxsdvajdc ion exchan gers arc fu nctionFof pTT th e
fo rmer reaching m oderately constant valucsaTpITabove aboutU aniJthc latter at
EO^£i2fyi}boin/5pi^tOH'lbT tlicTcTuil exchangc'cap'a^cilies, expm as mcq. per
g of dry resin, for a few typical resins arc: ZeegFit 225 (Na form), 4.5-5 ; Zerolit 226

term absorption is used whenes cr ions or oilier solutes arc t.ikcn up bv an ion exchanger. It does
imply any specific typc.s of forces responsible for this uptake.

170

ION EXCHANGE VII, 2

(H form), 9-10; Zerolit FF (Cl form), 4^0; Zerolit G (Cl fom), 4.0, ^e total
exchange capacity expressed as,raeq.cm“^ of the wet resin is about of the
meqrr'^'oTfEe dry resin. These figures are useful in estimating very appro jdmate^
the quantity ol resin required in a determination; an adequate excess must be
employed, since the ‘break through’ capacity is often much less than the total
capacity of the resin. In most cases a 100 per cent excess is satisfactory.

The exchange capacity of a cation exchange resin may be measured in the
l aboratory bv determinirig the numher of milligram equivalents of soSium i6n~
whic h are absorbed bv 1 g of the drv resin in the hydrogen fortn . Similarly, the
^chatige capacity of a strongly basic anion exchange resin is evaluated b y
measuring tlm amount of chloride ion taken up~ by j .g ol' dry resm in The
Sydroxide form..

Changing the ionic form. Some widely used resins. It is frequently necessary
to convert a resin completely from one ionic form to another. This should be done
after regeneration, if this is being practised to ‘clean’ the resin (e.g., if the ‘standard’
grade of ion exchanger is used). An excess of a suitable salt solution should be run
through a column of the resin. Ready conversion will occur if the ion to be
introduced into the resin has a higher, or only a slightly lower, affinity than that
actually on the resin. When replacing an ion of lower valency on the exchanger by
one of higher valency, the conversion is assisted by using a dilute solution of
replacing salt (preferably as low as 0.01 Af), while to substitute a higher valent ion
in the exchanger by one of lower valency, a comparatively concentrated solution
should be used (say, a IM solution).

Strongly acidic cation exchangers are usually supplied in the hydrogen or
sodium forms, and strongly basic anion exchangers in the chloride or hydroxide
forms; tlie chloride form is preferred to the free base form, since the latter readily
absorbs carbon dioxide from the atmosphere and becomes partly converted into'
the carbonate form. Weakly acidic cation exchangers are generally supplied in
the hydrogen form, while weakly basic anion exchange resins are available in the
hydroxide or chloride forms.

Strongly acidic cation exchangers ( po lystyrene sulphonic acid resins) —
Zerolit 225. Amberlite 120. etc. These re sms are u sually marketed in the sodium
Jorm,* and to convert them into the hydrogen form (w hich, it may be noted, are
' also available commercially) the following procedure may be used.

The resin, after regeneration (see Section Vn, 7) if the ‘standard’ grade is used,
is treated with 2M- or with 10 per cent hydrochloric acid: one bed volume of the
acid is passed through the column in 10-15 minutes. The effluent should then be
strongly acid to methyl orange indicator; if it is not, further acid must be used
(about three bed volumes may be required). The excess of acid is drained to
almost bed level and the remaining acid washed away with distilled or de-ionised
water, the volume required being about six times that of the bed. This operation
occupies about 20 minutes: it, is complete when the final 100 cm^ of, effluent
requires less than 1 cm^ of 0.02M sodium hydroxide to neutralise its acidity using
methyl orange as indicator. The resin can now be employed for the exchange of its
hydrogen ions for cations present in a given solution. Tests on the treated effluent
show that its acidity, due to the exchange, rises to a m’i.'F'num, which is

* The resin is supplied in moist condition, arid should not be 'allowed to ' ide fractur'-mav

occur after repeated drying arid re- wetting.- - . * '-’v •

VI!, 3 QUANTITATIVE INORGANIC ANALYSIS

maintained until the capacity is exhausted when the acidity of the treated solution
falls. Rcccncration is then necessary and is performed, after back-washing, with
2.4/ hydrochloric acid as before.

4\'caklv ac idic cation exch angers (polymcthylacrylic acid, e tc., rcsin.s)—
Zerolit 226, Amberhte3oScZ7i'l'cse resms arc lisualTy siippITcd in fTieltydrcrgcn
IbrnirThcy arc readily changed into the sodium form by treatment with Af-
sodium hydroxide; an increase in volume of80- 100 per cent may be expected, the
swelling is reversible and does not appear to cause any damage to the bead
structure. Below a pH of about 3.5, the hydrogen form exist.s almost entirely in the
little-ionised carboxylic acid form. Exchange with metal ions will occur ip
solution only when these arc associated in solution with anions of weak acids, i.c.,
pH values above about 4.

The exhausted resin is more easily regenerated than the .strongly acidic
exchangers: about 1.5 bed volumc.sof I4/-hydrochloricacid will usually suffice.

Strongly hasi c-jmiDn- Cxcliangcrs ( poly styrene quat ernar y amm onium
resins— Zerolit FF, Ar nberlite 400, etc. These resins arc usually supplicdlrTthe
chloruJiTfofm. hor coliversion into the hydroxide form, treatment with 14/-
sodium hydroxide is employed, the volume used depending upon the extent of
conversion desired; two bed volumes are satisfactory for most purposes. The
rinsing of the resin free from alkali should be done with dc-ionised water free from
carbon dioxide to avoid convening the resin into the carbonate form; about 2
dnr' of such water will suffice per lOOgofresin. An increase in volume of about 20
per cent occurs in the conversion of the resin from the chloride to the hydroxide
form.

Weakly basic an iDUx-XChan gcrs (polv stvrcDC-tcrtian'. ami ne resins)— Zerolit
G, Arnberlite 4,5, et c.,... These resins are generally .supplied in the free base
'tliydroxidcl form, 'rhe salt form may l>e prepared by treating the resin with about
four bed volumes of the appropriate acid (c.g., 1 4/-hydrochloric acid) and rinsing
with water to remove the excess of acid; the final effluent will not be exactly
neutral, since hydrolysis occurs slowly, resulting in slightly acidic effluents. As
with cation exchange, rpiantilaiivc anion exchange w ill occur only if the anion in
the resin has a lower aflinity for the resin than the anion to be exchanged in the
solution. When the resin is exhausted, regeneration can be accomplished by
treatment with cxce.ss of l.\/-sodium hydroxide, followed by washing with de-
ionised water until the effluent Is neutral. If ammonia solution i.s used for
regeneration the amount of washing required is reduced.

J^3. ION EXCHANGE CHROM.A'rOGRAPHV. If a mixtun^ailnvaj^
more dilTerent ca tions. A. B, et c., is passed throu gh an iQjixixmuumerQlmnn. and if
tlicquantities omicse ions arcsmall compITrod with the total capacity of the
column for ions, then it may bcj K?^sibi c to recover the absorbed ions separately
.and consecutively byjusinga suitable rcii^rating (oradini nel solution. iT cIitIdn
AJsjw ld more firmly by th e cxaffiTTgrreiffirnian ca tionSf^Hhe-B-prcscm-will
flow offi _ofilic_lr oitom oTTlic column bcforc.any.of A. jsjibcrated, provided that
theemumn is long enough and other experimental factors arc favourable for the
particular separation. This separation techniqu e is .sometirnesaxalledionexehange
chromatography. Its most spectacular success has~bceri the separation of the
.Ja nthanoids and als o o| ot her cations of verv .similar properties (c.c.. HI and Zr,'
.Nba ndTliTNa and K)~ ' ^

TTtc~procTrss of removing absorbed ions is known as elution, the smTition

172

ION EXCHANGE VH, 3

Volume of eluate. cm^

Volume of eluate, cm'

'g.vn,i

employed for elution is termed the and the .so lution r esulting from elution

Ts'called the eluate. The liquid enterilfglhe ion exchange colum n may be termed
lEeuffluent and the liquid leavi ng the column is conve nienW^all e d the effluent, l l'
^solution of a suitable eluant is pass^ through a column charged with an ion A'
the course of the reaction may be followediiy analysing continuously the effluent
solution. If the concentration of A in successive portions of the eluate is plotted ’
against the volume of the eluate, an elution curve is obtained such as is shown in
Fig. VII, 1. It will be seen that practically all the A is contained in a certain volume
of liquid and also that the concentration of A passes through a maximum. ,

If the ion exchange column is loaded with several ions of similar charge, B, C,
etc., elution curves may be obtained for each ion by the use of appropriate eluants.

If the elutio n curves are sufficiently far apart, as in Fig. VII, 2. a quantitative
sepSatTon is possible: only an incomplete separation is obtained if the elution
curves overlap. Ideally the curves should approach a Ciaussian (norti
distribution (Section TV. 8 ) and e xcessive departure f rom th is distribution may
indicate faulty technique and/or column operating conditions.

Jhe rate at which two constituents separa te in the column is determined by the
ratio of the two corresponding distribution
coefficient is given by the equation

amount of solute on resin
weight of resin, g

amount of solute in solution
volume of solution, cm^

The distribution coefficient can be determined by batch experiments in which a
smalt known quantity of resin is shaken with a solution containing a known
concentration of the solute, followed by analysis of the two phases after
equilibrium has been attained . The separation factor, a. is used as a measure of the
chromatographic separation po s -sthie and is given by the eqi^ion .

p, w‘

where and are the distribution coefficients of the two constituents. The
4 ^ter the dev i atibiTof a from unity the easier will be the se paration, h'of normal
laEoI'aiory practice, a uselul guide is that quantitative lena ra'finn cbrmirl hp
achieved if « is above 1.2 or less tha n 0.8.

important relationsFip exists between the weight distribution coefficient
and the volume of eluant required to reach the maximum concentration of

173

\1I, 3 QUANTITATIVE INORGANIC ANALYSIS

an eluted ion in the cfTluent. Tliis is given by the equation.

where V is the volume of liquid in the interstices between the individual resin
beads. If the latter arc spheres of uniform size and close-packed in the column, V„
is approximately 0.4 of thcj otal bed vo lume^ The void fraction VJV^, of the
column may, however, be detcrrmneiTcxperimenially or calculated from density
data (Ref. 2).

The volume distribution cocRicient is also a useful parameter for chromato-
graphic calculations and is defined as

amount of ion in 1 cm^ of resin bed
■■ ' amount of ion in 1 cm^ of interstitial volume

It is related to the weight distribution cocfllcicnt by

D, - KJi

where// is the void fraction of the settled column.
It IS also related to by the equation

It should be remembered that the relationships given above arc strictly
applicable only when the loading of the column is less than 5 per cent of its
capacity.

The application of these parameters may be illustrated by the following
example.

Example. A mixture of ca. 0.05 meq each of chloride and bromide ions is to
be separated on an anion exchange column of length 10 cm and 1 cm* cross
section, using 0.035;\/ potassium nitrate as the eluant. The distribution
cocllicients (R,,) for the chloride and bromide ions rc.spcctivcly arc 29 and 65.

Separation factor a =

2.24

This value indicates that a satisfactory separation could be achieved, and this is
confirmed by calculation of the values for the appearance of the cliloridc and

bromide peaks.

From the column dimensions, the bed volume is

I't = lOcm X l.Ocm^ = 10.0 cm^
and the void volume {ns.suming // = 0.4) is
l'„ — 0.4 X 10.0 cm^ = 4 cni^

Hence for the chloride peak,

Ka. = A'ji; + 1; - (29 X 4) + 4 = 1 20 cm^
and for bromide,

= (65 X 4) + 4 = 264 cm^

The relatively large values of indicate, however, that the separation will be

174

ION EXCHANGE VH, 4/5

lengthy and the elution bands broad, particularly for the bromide band. The use
of a more concentrated solution of eluant significantly reduces the values of
and the elution bands become much sharper. Thus the distribution coefficient for
bromide using a 0.35M potassium nitrate solution is 6.5 and using the same
column, = (6.5 x 4) + 4 = 30 cm^.

In many cases the efficient ' separation of a mixture by ion exchange
chromatography requires that the eluant concentration be changed during the
course of the elution. This may be done in a stepwise manner or by a continuous
change in concentration as in gradient elution; the latter procedure can be carried
out using simple laboratory equipment. A comprehensive discussion of the
technique and of gradient elution devices is given in the review by L. R. Snyder
•(Ref. 3). .■ '■■■•- ■ ■

The scope of separations by ion exchange chromatography may be extended
b.y using for fixation or for elution a solution capable of complexing the ions .
exchanged. The formation of complexes may assist separations by diminishing
the concentrations of free ions, and also by producing complexes of different .
stabilities, thus leading to significantly different behaviour with selected eluants.

The results of ion exchange separations may be influenced by varying the pH,
the solvent or eluant, the temperature, the nature of the ion exchange resin, the
particle size, the rate of flow of eluant, and the length of the column.

vn, 4. ION EXCHANGE IN ORGANIC AND AQUEOUS-ORGANIC
SOLVENTS. Investigations in aqueous systems have established many of the
fundamental principles of ion exchange as well as providing useful applications.
The scope of the ion exchange process has, however, been extended during the
last decade or so by the use of both organic and mixed aqueous-organic solvent
systems (Ref. 4 and 5).

The organic solvents generally used are oxo-compounds of the alcohol, ketone
and carboxylic acid types, generally having dielectric constants below 40. Cations
and anions should, therefore, pair more strongly in such solvent systems than in
water and this factor may in itself be expected to alter select! vities for the resin. In
addition to influencing these purely electrostatic forces, the presence of the
organic solvent may enhance the tendency of a cation to complex with anionic or
other ligands thus modifying its ion exchange behaviour. In mixed aqueous-
organic solvents the magnitude of such effects will clearly be dependent on the
proportion of organic solvent present. • -

As already indicated, ion exchange resins are osmotic systems which swell
owing to solvent being drawn into the resin. Where mixed solvent systems are '
used the possibility of preferential osmosis occurs and it has been shown that
strongly acid cation and strongly basic anion ’ resin phases tend to ' be
predominantly aqueous with the ambient solution predominantly organic. This '
effect (preferential water sorption by the resin) increases as the dielectric constant
of the organic solvent decreases.

An interesting consequence of selective sorption is that conditions for partition
chromatography arise which may enhance the normal ion exchange separation
factors. This aspect has been utilised by Korkisch (Ref. 6) for separation of
inorganic ions by the so-called ‘Combined Ion Exchange-Solvent Extraction
Method’ (CISE), and is illustrated by experiment VII, 16.

CHELATING lON. EXCHANGE RESINS. The use of complexing

175

vn, 5 QUANTITATIVE INORGANIC ANALYSIS

agents in solution in order to enhance llic cflicicncy of separation of cation
mixtures (e.g. lanthanoids) using conve ntionaTcaUQn or anToirrcxchancc resins^
lI^'estaBlisTiHrAFrrn^^^ nTo3c of application of complex formation is,

however, the u'se of chelating resins which arc ion exchangers in which various
chelating groups (c.g., dimethyiglyoximc and iminodiacetic acid) have been
incorporated and arc attached to the resin matrix.

An important feature of chelating ion c.xchangcrs is th e greater selectivity
which they offer compared with the conventional type of ion exchang er. The
afiinity of a particular mctaITbliT6FircTfiaiircKcT:rtifr^csnrcrepchds mainly on
the nature of the chelating group, and the selective behaviour of the resin is
largely based on the different stabilities of the metal complexes formed on the
resin under various pff conditions. It may be noted that the bi nding energy in
Uicsc resins is of the order of 60-105 kJ mole'*, whereas jmjtjdin ary ion
"exchanger s the stn TiTfflTuflliirelecirosta'tic'bin'ding is only about 8-! 3 kj mole ~
TliFcxchangc process in a chelating rcsT7rTr'gcnct^I\'~s TowanHal ^
ordinaiyTypTbT^'xcRanprr tlie rate apparently being co ntroll ed bya~partlHc~
^nTtls Toh r ng^uinTsm.~7r~~~ ' " ' ~

~~*According to Gregor ci <il. (R.tT. 7) the following prp perlics.are required for a
chelating agent which is to be incorporated as a functioTial group into an ion
exchange resin;

1 . the chelating agent should yield, either alone or witii a cro.ss-linfcing substance,
a resin gel oLsnIheient stability or be capable of incorporation into a polymer
matrix;

2. the chelating group must have sufficient chemical stability, so that during the
synthesis of the resin its functional structure is not changed by polymerisation
or any other reaction :

3. the steric structure of the chelating group .should be compact so that the
formation of the chelate rings with cations will not he liindcrcd by the resin
matrix;

4. the specific arrangements of the ligand groups should be preserved in the resin.
This is particularly necessary since the complexing agents forming sufliciently
stable complexes arc usually at least tridcntatc.

The,sc considerations indicate that many chelating agents could not be
incorporated into a resin without loss of (heir selective complexing abilities.
Ligands which do not form 1 : 1 complexes (c.g., 8-quinoIinoI) would be
unsuitable, as also would molecules such as EDTA, which arc insuflicicntly
compact. In the latter ease, it is improbable that the chelate configurations
occurring in aqueous solution could be maintained in a cro-ss-linkcd polymer.
The closely related iminodiacetic acid group does, however, meet the
requirements described, being compact and forming 1;I complexes with metal
cations.

ION EXCHANGE VH, 6

Although chelating resins containing various ligand donor atoms have been
synthesised, the iminodiacetic acid resins (N and O donor atoms) undoubtedly -
form the largest group (Refs. 8 and 9). The resin based on iminodiacetic acid in a
styrene-divinylbenzene matrix is available..commercially under the trade names
of Dowex Chelating Resin A-1 and Chelex 100, and its chemical and physical
properties have been fully investigated. >

The starting material for the synthesis of this chelating resin is '
chloromethylated styrene-divinylbenzene which undergoes an amination re-
action and- is then treated with monochloracetic acid :

-CH— CHj—

'CH,C1

NH,

-CH— CHa—

'CHaNHa

— CH— CHa—

CH,ClCOOH

CH,COOH

The selectivity of this type of exchange resin is illustrated by Chelex 100 which
shows unusually high preference for copper, iron and other heavy metals (i.e.,
metals which form complexes having high stability constants with this type' of
ligand) over such cations as sodium, potassium and calcium ; it is also much more
selective for the alkaline earths than for the alkali metal cations. The resin’s high
affinity for these ions makes it very useful for removing, concentrating or
analysing traces of them in solutions, even when large amounts of sodium and
potassium are present. . , ,

I, 6. LIQUID ION EXCHANGERS. The ion exchange processes
involving exchange resins occur between a solid and liquid phase whereas in the
case of liquid ion exchangers the process takes place between two immiscible
solutions. Li quid ion exchanger s consist of high molecu l ar weight acids and bases
which possess low solubility in~w^r and high solubility m water-immiscible '
gplvents. I hus a solution of a base insoluble m water, in a solvent which is water-
immiscible, cari~Ee used as an anion exchanger; similarly a solution of an acid
insoluble in water can act as a cation exchanger for ions in aqueous solution.^
comprehensive list of liquid ion exchangers has been given by Coleman et al. (R^f.'

Th ^liquid am^ exchangers at present available are based largely on primary.

tertiary aliphatic amines, e.g., the exchangers , Amberlite
- [N-dodecenyl(trialkylmethvl)amine] and Amberlite L'A'. 2 TN^

l auryl(trialkylmethvl)amineiy b oth secondary amines. These anion exchang e
Jig pids are best employed as solutions (ca. 2.5 to 12.5% v/v l in an inert org ^c
s^enrsuch.as. be.nzene. toluene, kerosene, petroleum ether. c^lohexane,~octahB^ .

The liquid exchangers Amberlit e LA.l and I.A.2 may h p used to frmnvf nrids
■from SOlutir>n ^

R'R''NH + HX — » R-:R"NH2X

£LlIL£ j3lt form for yarious ion exchange processes

R'R"NH2C1+' NaNOj — > R'R"NH 2 N 03 + Naa
Example^s_o niqui d catiqp^g^angers ar e alkyl and diah^l phosphoric acids.

177

VII, 7 QUANTITATIVE INORGANIC ANALYSIS

alkyl suloho nic acids and carbo xylic acids, although only two appear to have
been used to " any extent, vE,~ jli-(2^tliylhexyl)orthopliosphoric acid a nd
. dinoiiylnap hlha lcne sulphonic n citL

The opcratron ofliquid ion exchangers involves tlu LSelectivc transfer of a solute
between an aqueous nhas cand an immiscible orcanic phase containinel liciiaiiid
'~cxHi?m'cc'nr.''Tlfus TiipirinHcallar^u^ amin cijS^-aad-iiolutlQ^-icId .large
“TSticnsraTrablcofforming extractable species (c.g., ion pairs) with various anions.
ThtrTttClui^TiirenTpIoycdTdrsepafationrusih'f ion 'exchangers is thus

identical to that used in solvent extraction separations and these c.xchangcrs thus
offer many of the advantages of both ion exchange and solvent extraction. There
arc, however, certain diflicultics and disadvantages associated with their u.se
which it is important to appreciate in order to make clTective use ofliquid ion
exchangers. .

Probably the cliic fjdnucultv which arises is that due to the formation of
e miil. sjp ns belwee n the organic and aqueous phases. This makes separ ation o f the~
^ili jerrfrTFicult a mLsp.mcimres inrpossIlrlcrTns~Bearlvjmp ortanl to select liquid
J pxclianpcrs having lo w surface activity a nd to u.sccon^itibns whiclnyiilj™
t lic Ibri^tion ol '.s^blc cmiiTsions (sec Section VI, 6).

" Another disadvantage inTlieTisc ofliquid ion exchangers is thatjt i s freq uentlv
necessary to back-e xtract th e require d species front-the orc anic p hase into an‘
aoucous phase prior to completing t he determ inati on. T he organic phase mayl
however, sometimes be used directly forUclcrmination of the extracted species, in
particular by aspirating directly into a flame and estimating extracted metal ions
by flame photometry or atomic absorption spectroscopy.

The extraction of metals by liquid amines has been widely investigated and
depends on the formation of anionic complexes of the metals in aqueous solution.
Such applications are illustrated by the use of Ambcrlitc LA,1 for extraction of
zirconium and hafnium from hydrochloric acid solutions, and the use ofliquid
amines for extraction of uninium from sulphuric acid solutions (Refs. 1 1 and 12).

Exhausted liquid ion exchangers may be regenerated in an analogous manner
to ion exchange resins, e.g., Ambcrlitc LA. I saturated with nitrate ions can be
converted to the chloride form by treatment with excess sodium chloride
solution.

APPLICATIONS IN ANALYTICAL CHEMISTRY

VII, 7. EXPERIMENTAL TECHNIQUES. The simplest apparatus for ion
exchange work in analysis consists ofa burette providcri wilinrgtass-wuufpliig or
•£i!li£Ed-£lassjiisc^(porosity 0 or 1 ) aTTTirioTwTna77rnoTIra-7an
shown in Fig. VIl, 3, {a): the ion c.xchangc resin is supported on a glass-wool plug
or sintcrcd-glass disc. A cl ass-wool pad mav Ire placeclat the top' of the bed of'
■*^cgiD_and_lh.e_eluliaii-age nt is addedfWmra~t:ip Tu nnel supportcdlitlbTclfic -
^lumn. The siphon overflowTubc, attached to the cTiTuimrb^lTslibfTtenglh'bf
rubber or PVC tubing, ensures that the level of the liquid docs not fall below the
top of the resin bed, so that the latter is always wholly immensed in the liquid. Tire
ratio of the height of the column to the diameter is not very critical but is usually
10 or 20 to 1. Another form of column is depicted in Fig. VII, 3, (/))(not drawn to
scale): a convenient size is 30 cm long, the lower portion of about 10.mm and the

178

vj/ 0

ION EXCHANGE VH, 7

upper portioii of about 25 mm
internal diameter. A commercially
available column, fitted with
ground-glass joints, is illustrated
in Fig. VII, 3, (c).

The ion exchange resin should
be of small particle size, so as to
provide a large surface of contact;
it should, however, not be so fine
as to produce a very slow flow
rate. For most analytical work
50-100-mesh or 100-200-mesh
materials are satisfactory. In all
cases the diameter of the resin
bead should be less than one-tenth
‘oTThat of the column. Resins of
medium and high cross-linking
rarely show any further changes in
volume, and only if subjected to
large changes of ionic strength will
any appreciable volume change
occui(^esins of lo^^cross-linking may change in volume appreciably even with
small variations of ionic strength, and this may result in channelling and possible
blocking of the column; these effects limit the use of these material^To obtain
satisfactory separations, it is essential that the solptions shpuld passThrough the
column in a uniform ma nnen The resin particles should be packed uniibrrnly in •
the_column: the resin bed should be free from air bubbles so that there is no '

channelling! - ’ ^

To prepare' a well-packed column, a supply of exchange, resin of narrow size
range is desirable..^An-i on exchange resin swells if the dry solid is immersed i n
jvater; no attempt should therefore be made to set up a column by pouring the dry
resin into a tube and then adding water, since the expansion will probably shatter
the tube. The resin should he stirred with water in an open beaker for severa l
jninutes, any fi ne parti cles removed hv decantation, and the resin slurry
t ransferred portionwise to the tube previously filled with water. TIieTuEe may be
i§B Egd gently to prevent the formation of air bubbles . To ensure the removal oT
entrained air bubbles, of any remaining fine particles, and also to ensure an even
^MabmiQiLQl xesin granules, it is advisable to ‘backwash’ the resin column belofe
Jise, i.e., a stream of good quality distilled water or of de-ionised water is nm up
through the bed from the bottom at a sufficient flow rate to loosen and suspend
the exchanger granules. The enlarged upper portion of the exchange tube shown
in Fig. VII, 3, b or c, will hold the resin suspension during washing. If a tube of
uniform bore is used the volume of resin employed must be suitably adjusted or
else a tube attached by a rubber bung to the top of the column; the tube dips into
an open filter flask, the side arm of which acts as the overflow and is connected by
rubber tubing to waste. When the wash water is clear the flow of water is stopped
and the resin is allowed to settle in the tube. The excess of water is drained off; the
level must neve r fall below the surfa ce o f th e resi n, or else channelling will
•^^ilCMJvith coniequent incomplete contact between the resin andlblutionsTKed — -
in subsequent operations. The apparatus with a side ■ (Fig. VII, 3, a) has

179

8 QUANTITATIVE INORGANIC ANALYSIS

an advantage in this respect in that the resin will not run dry even if left
unattended, since the outlet is above the surface of the resin.

Ion exchange resins (standard grades) as received from the manufacturers may
contain unwanted ionic impurities and sometimes traces of water-soluble
intermediates or incompletely polymerised material; these must be washed out
before use. This is best done by passing 2Af-hydrochlorfc acid and 2Af-sodium
hydroxide alternately through the column, with distilled-watcr rinsings in
between, and then washing with water until the efiluent is neutral and salt free.
‘Analytical grade* and/or ‘chromatographic grade’ ion exchange resins that have
undergone this preliminary washing arc available commercially.

For analytical work the exchange resin of ‘analyticar grade (Ambcrlite) or of
‘chromatographic’ grade (Permutit; Ambcrlite, etc.) of a particle size of 100-200
mesh is preferred. However, for student work, the ‘standard’ grade of resin of
50-100 or 15-50 mesh, which is less expensive, is generally satisfactory. The
‘standard’ grade of resin must, however, be conditioned before use. Cation
exchange resins must be soaked in a beaker in about twice the volume of2A/-
hydrochloric acid for .10 -60 minutes with occasional stirring; the fine particles
arc removed by dccanttition or by back-washing in a column with distilled or de-
ionised water until the supernatant liquid is clear. Anion exchange resins may be
washed with water in a beaker until the colour of the decanted wash liquid
reaches a minimum intensity; they may then be transferred to a svidc glass
column and cycled between lAf-hydrochloric acid and !,\f-a!kaii. Sodium
hydroxide is used for strongly basic resins, and amntonia (preferably) or sodium
carbonate for weakly basic resins. For all resins the final treatment .should be with
a solution leading to the resin in the desired ionic form.

A 50-cm-' or 1 (X)-cm’ burette, with Pyrex gla-ss-wool plug or sintered-glass disc
at the lower end, cm generally be used for the determinations dc.scribcd below:
alternatively, the column with .side arm (Fig. VII. 1, u) is equally convenient in
practice for student use. Reference will be made to the Permutit resins; the
equivalent Amberlitc or other resin (sec Table VJI, 1 in Section VII, I) may of
course be used.

Vrr. 8. DETFRMINATION OF THE CAPACITY OF AN ION EX-
CHANGE RESIN (COLUMN METHOD). Cation exchange resin. Dry the .
purified resin (c.g., Zcrolit 225 in the hydrogen form) by placing it in an,
evaporating disli, cover with a clock glass supported on two glass rods to provide,
protection from dirst while giving access to the nir, and leave in a warm place (25-
35 ‘'C) until the resin is completely free-running (2-3 days). Tlic capacity of the
resulting resin remains constant over a long period if kept in a closed bottle.
Drying at higher temperatures (say. 100 C) is not recommended, owing to
possible fracture of the resin beads.

Partly fill a small column, 15 cm x I cm (Fig. VII, 3, u) with distilled water,
taking care to displace any trapped air from beneath the sintered-glass disc.
Weigh out accurately about 0.5 g of the air-dried resin in a glass scoop and
transfer it with the aid of a small camel-hair brush through a dry funnel into the
column. Add sufficient distilled water to cover the resin. Dislodge any air bubbles
that stick to the resin beads by applying an intermittent pressure to the rubber
tubing, thus causing the level of the liquid in the column to rise and fall slightly.
Adjust the level of the outlet tube so that the liquid in the column will drain to a-
level about 1 cm above the resin beads.

180

ION EXCHANGE VH, 9

Fill a 250-cm^ separatory funnel with ca. 0.25M-sodium sulphate solution.
Allow this solution to drip into the column at a rate of about 2 cm^ per. minute,
and collect the effluent in a 500-cm^ conical flask. When all the solution has
passed through the column, titrate the effluent with standard O.lM-sodium
hydroxide using phenolphthalein as indicator.

The reaction may be represented as;

2R"H+ +2Na+ ^2R"Na+ +2H+

and proceeds to completion because of the large excess and large volume of
sodium sulphate solution passed through the column.

The capacity of the resin in milli-equivalents per gram is given by av/W, where
a is the molarity of the sodium hydroxide solution, v is the volume in cm^, and W
is the weight (g) of the resin.

Anion exchange resin. Proceed as in the previous experiment using 1.0 g,
accurately weighed, of the air-dried strongly basic anion exchanger (e.g., Zerolit
FF, chloride form). Fill the 250-cm^ separatory funnel with ca. 0.25M-sodium
nitrate solution, and allow this solution to drop into the column at the rate of
about 2 cm^ per minute. Collect the effluent in a 500-cm^ conical flask, and titrate
with standard O.lM-silver nitrate using potassium chromate as indicator.

The reaction which occurs may be written as;

R-'Cl--t-N 03 -:^R+N 03 "-fCr

The capacity of the resin expressed as milli-equivalents per gram is given by bv/W,
where o cm^ of bM AgNOj are required by ^ g of the resin.

Vn, 9. SEPARATION OF ZINC AND MAGNESIUM ON AN ANION
EXCHANGER. Theory. Several metal ions (e.g., those of Fe, AI, Zn, Co, Mn,
etc.) can be absorbed from hydrochloric acid solutions on anion exchange resins
owing to the formation of negatively charged chloro complexes. Each metal is
absorbed over a well-defined range of pH, and this property can be used as the
basis of a method of separation. Zinc is absorbed from 2M-acid, while
magnesium (and aluminium) are not; thus by passing a mixture of zinc and
magnesium through a column of anion exchange resin a separation is effected.
The zinc is subsequently eluted with dilute nitric acid.

Procedure. Prepare a column of the anion exchange resin using about 15 g
of Zerolit FF in the chloride form (Section VII, 7). The column should be made up
in 2M hydrochloric acid.

Prepare standard zinc (about 2.5 mg Zn/cm^) and magnesium (about 1.5 mg
Mg/cm ) ion solutions by dissolving accurately weighed quantities of A.R. zinc
shot and magnesium (for Grignard reaction) in 2M-hydrochloric acid and
diluting each to volume in a 250-cm^ graduated flask. Pipette 10.0 cm^ of the zinc
ion solution and 10.0 cm^ of the magnesium ion solution into a small separatory
unnel supported in the top of the ion exchange column, and mix the, solutions.
How the mixed solution to flow through the column at a rate of about 5 cm^ per
minute. Wash the funnel and column with 50 cm^ of 2M-hydrochloric acid: do
Ihs level of the liquid to fall below the top of the resin column. Collect
n the effluent in a conical flask ; this contains all the magnesiurh. Now change the
receiver. Elute the zinc with 30 cm^ of water, followed by 80 cm^ of ca. 0,25M-
' acid. Determine the magnesium and the zinc in the respective eluates by
eu ralisation with sodium hydroxide solution, followed by titration with

181

VII, 10 QUANTITATIVE INORGANIC ANALYSIS

Standard EDTA solution using a buITcr solution of pH - 10 and Solochrome
Black indicator (Sections X, 64, 67).

The following results were obtained in a typical experiment:

Weight of zinc taken = 25.62 mg, found = 25.60 mg
Weight of magnesium taken = 14.95 mg, found = 14.89 mg

Magnesium may conveniently be determined by atomic absorption spectro-
scopy (Section XXir. 22) if a smaller amount (ca. 4 mg) is used for the separation.
Collect the magnesium cfllucnt in a 1 dm^ graduated flask, dilute to the mark with
dc-ionised water and aspirate the solution into the flame of an atomic absorption
spcctronietcr. Calibrate the instrument using standard magnesium solutions
covering the range 2 to 8 p.p.m.

VII, 10. .SEPARATION OF CHLORIDE AND BROMIDE ON AN ANION
EXCHANGER. Theory, The anion c.vchange resin, originally in the chloride
form, is converted into (he nitrate form by washing with sodium nitrate solution.
A concentrated solution of the chloride and bromide mixture is introduced at the
top of the column. The halide ions exchange nipidly with the nitrate ions in the
resin, forming a band at the top of the column. Chloride ion is more rapidly eluted
from this band than bromide ion by sodium nitrate solution, so that a separation
is possible. The progress of elution of the luilides is followed by titrating fractions
of the efllucnts with standard silver nitrate solution.

Procedure. Prepare an anion exchange column (Section MI, 7) using
about 40 g of Zerolit FI' (chloride form). The ion exchange tube may be 16 cm
long and about 12 mm internal diameter. Wash the column with 0.6.\/-.sodium
nitrate until the cfllucnt contains no chloride ion (silver nitrate test) and then
wash with 50 cm^ of0.3.\/-sodiuni nitrate.

Weigh out accurately about 0. lOg of A.R. sodium chloride and about 0.20 g of
A.R. potassium bromide, dissolve in about 2,0 cm’ of water and transfer
quantitatively to the top of the column with (he aid of0.3Af-sodium nitrate. Pass
0,3.\/-sodium nitrate through thecolumn at a flow rate of about 1 cut’ pcrminule
and collect the effluent in 10-cm^ friiciions. Transfer each fraction in turn to a
conical flask, dilute with an equal volume of w;iter, add 2 drops of 0.2A/-
potassiiim chromate solution and titrate with standard 0.02A/-silver nitrate.

Before commencing the elution titnite lO.O cm^ of the 0.3A/-sodium nitrate
with the standard siKcr nitrate solution, and retain the product of the blank
titration for comparing with the colour in the actmil titrations of the eluates.
When the titre of (he cluale falls almost to zero (i.c., nearly equal to the blank
titration) ctt. 150 cm ' of cfliuent elute the column with d.6A/-sodium nitrate.
Titrate as before until no more bromide is detected (litre almost zero), A new
blank titration must be made with 10.0 enf’ of the0,6A/-sOdium nitrate.

Plot a graph of the total cfliuent collected against the concentration of halide in
each fraction (millimois per litre). The sum of the titres using 0.3 A/-sodium nitrate
eluant (less blank for each titration) corresponds to the chloride, and the parallel
figure with 0,6A/-sodium nitrate corresponds to the bromide recovery.

A typical experiment gave the following results;

Weight of sodium chloride u.scd = 0,1012 g s 61.37 mg Cl"

Weight of potassium bromide u.scd 0.1934g sr 129.87 mg Br”
Concentration of silver nitrate solution = 0.01 936A/

182

ION EXCHANGE VH,. 11/12

Cr : total titres (less blanks) = 89.54 cm^. = 61.47 mg

Br," : total titres (less blanks) = 83.65 cm^ = 129.4 mg ^

Vn, 11. DETERMINATION OF THE TOTAL CATION CONCENTRA-
TION IN WATER. . Theory. The following procedure is a rapid one for the
determination of the total cations present in water, particularly that used for
industrial ion exchange plant, but may be used for. all samples of water, including
tap-wafer. When water containing dissolved ionised solids is passed through a
cation exchanger in the hydrogen- form all cations are removed and replaced by
hydrogen ions. By this means any alkalinity present in the water is destroyed, and
the neutral salts present in solution are converted into the corresponding mineral
acids. The effluent is titrated with 0.02M-sodium hydroxide using screened
methyl orange as indicator.

Procedure. Prepare a 25-30-cm column of Zerolit 225 in a 14-16-mm
chromatographic tube (Section VII, 7). Pass 250 cm^ of 2M-hydrochloric acid
through the tube during about 30 minutes; rinse the column with distilled water
until the effluent is just alkaline to screened methyl orange or until a 10-cm^
portion of the effluent does not require more than one drop of 0.02M-sodium
hydroxide to give an alkaline reaction to bromothymol blue indicator. The resin
is now ready for use: the level of the water should never be permitted to drop
below the upper surface of the resin in the column. Pass 50.0 cm^ of the sample of
water under test through the column at a rate of 3-4 cm^ per minute, and discard
the effluent. Now pass two lOO.O-cm^ portions through the column at the same
rate, collect the effluents separately, and titrate each with standard 0.02M-sodium
hydroxide using screened methyl orange as indicator. After the determination has
been completed, pass 100-150 cm^ of distilled or de-ionised water through the
column.

From the results of the titration calculate the milli-equivalents of calcium
present in the water. It may be expressed, if desired, as the equivalent mineral
acidity (E.M.A.) in terms of mg CaCOj per dm^ of water (i.e., parts per million of
CaCOj). In general, if the titre is A cm^ of sodium hydroxide of molarity B for an
aliquot volume of V cm^, the E.M.A. is given by (AB x 50 x 1000)/F.

Commercial samples of water are frequently- alkaline due to the presence, of
hydrogen carbonates, carbonates, or hydroxides. The alkalinity is determined by
titrating a 100.0-cm^ sample with 0.02Af-hydrochloric acid using screened
methyl orange as indicator (or to a pH of 3.8). To obtain the total cation content
m terms of CaCOj, the total methyl orange alkalinity is added to the E.M.A.

12. SEPARATION OF COBALT AND NICKEL ON AN ANION
XCHANGER. Theory. The separation is based upon the fact that cobalt,
ut not nickel, forms a monovalent complex anion (probably [CoClj]") in 9M- ,
ydrochloric acid, and this anion is rapidly extracted from the solution by a
s rongly basic anion exchanger, such as Zerolit FF. The nickel is not retained by
e resin, presumably because of the instability of the anionic chloro complex,
n can be washed out of the column with 9M-hydrochloric acid. Upoii washing
e column with water, the cobalt complex is decomposed and passes out in the
cobalt(II) chloride. The nickel and cobalt in the respective effluents
y e determined, after evaporation of the excess of.hydrochloric acid, by

titration with EDTA. > x

Reagents. Anion exchange column. Prepare an anion exchange column

183

VII, 12 QUANTITATIVE INORGANIC ANALYSIS

using 25-30 g of Zerolit FF (chloride form). Mix the resin rvith about 100 cm^ of
water in a measuring cylinder and shake for a few minutes, decant the ]ic]uid as
soon as the larger particles have settled. The volume of the resin should be about
25 cm^. Stir the resin with distilled water, allow to settle, and deatnt the
supernatant liquidrrcpcatthcprocessuntil thc.supcrnatant liquid isclcar.Transfer
the resin slurry to a burette containing a plug of glass wool until a column of well-
packed resin about 22 cm long is obtained; alternatively, use an ion exchange
tube (see Fig. Vll, 3, a). Wash the resin in the column once wth water. Do not
allow the level of the liquid in the column to fall below the upper surface of the
resin : the level should preferably be about 1 cm above it.

Cobalt-ion solution. Di.ssolvc 5.0 g A.R. hydrated cobnlt{Il) chloride in 9M-
hydrochloric acid and dilute to 250cm-' with 9.M-hydrochloricacid.

Nickcl-ion .solution. Dissolve 2.5 g pure nickel carbonate in 9M-
hydrochloricacid and dilute to 250-cm^ with 9Af-hydrochloricacid.

Procedure. Pass 50 cm’ of 9A/-hydrochioric acid through the column and
drain to almost bed level. Mix 10.0 cm’ of each of the cobalt and nickcl-ion
solutions in a small beaker, transfer 10.0cm’ of the mi.vcd .solution with the aid of
a pipette to the top of the resin column, and lower this solution to the upper part
of the column with a little 9A/-hydrochloric acid. Pass 100 cm’ of 9Af-
hydrocWoric acid through the column in order to elute the nickel; collect the
cluatc in a 400-cm’ beaker. Concentrate the eluatc to a small volume on a wire
gauze (FUME CUPBOARD!) in order to remove the excess of acid. Neutralise
the resulting solution with A.R. potassium hydroxide, dilute to 100 cm’ with
distilled water, add 10 cm’ of buffer .solution (prepared by mixing equal volumes
of 1 A/-NH4CI and 1 A/-aqucous ammonia), about 15 drops of Bromopyrogallol
Red indicator .solution* and titrate with standard 0.02Af-EDTA until the colour
changes from blue to wine red (see Section X, 58(/))}. Perform ti similar titration
with 5.00 cm’ of the original nickcl-ion solution.

After the nickel has been eluted from the column, pass 150 cm’ of water at the
rate of 4~5 cm’ per minute through the resin to decompose the anionic cobalt
chloro complex, and collect the cfllucnt in a small beaker. Concentrate the
efiluent to a small volume, partly neutralise with A.R. potassium hydroxide, and
adjust the pH of tlic solution to about (> by the addition of powdered hcxaminc.
Add a few milligrams of Xylcnol Orange indicator,! warm to about 60 X, and
then titrate with standard 0.02Af-EDTA (slowly near the end-point) until the
colour changes from red to orange-yellow. Perform a similar titration upon 5.00
cm’ of the original cobalt-ion solution.

Compare the amount.s of nickel and cobalt recovered with those actually used.

Some typical results arc given below.

5.00 cm’ of the original nickcl-ion solution required 20.15 cm’ of 0.02A/-
EDTA. The nickcl-ion solution recovered after passage through the column
required 20.05 cm’ of 0,()2Af-EDTA.

5.00 cm’ of the original cobalt-ion solution required 21.45 cm’ of 0.02A/-
EDTA. The cobalt-ion solution recovered from the column required 21.35 cm’ of
0.02Af-EDTA.

* I'ull dciiiils of this intlic.itor.irc j'lvcn in Section X, 2H.
t Tin's iinJicalor is used ixs a solid mixture; details ate given in .Section .X, 28.

184

ION EXCHANGE VH, 13

VII, 13. SEPARATION OF CADMIUM AND ZINC ON AN ANION
EXCHANGER. Theory. Cadmium and zinc form negatively charged chloro
complexes which are absorbed by a strongly basic anion exchange resin, such as
Zerolit FF. The maximum absorption of cadmium and zinc is obtained in 0. 12M-
hydrochloric acid containing 100 g of sodium chloride per dm^. The zinc is eluted
quantitatively by a 2M-sodium hydroxide solution containing 20 g of sodium
chloride per dm^, while the cadmium is retained on the resin. Finally, the
cadmium is eluted with IM-nitric acid. The zinc and cadmium in their respective
effluents may be determined by titration with standard EDTA.

Elements such as Fe(III), Mn, Al, Bi, Ni, Co, Cr, Cu, Ti, the alkaline-earth
metals, and the lanthanoids are not absorbed on the resin in the HGl-NaCl
medium.

Reagents. Anion exchange column. Prepare an anion exchange column
using 25-30 g of Zerolit FF (chloride form) following the experimental details
given in Section Vn, 12. Allow the resin to settle in 0.5M-hydrochloric acid.
Transfer the resin slurry to the column: after settling, the resin column should be
about 20 cm in length if a 50-cm^ burette is used.

Reagent I. This consists of 0.12M-hydrochloric acid containing 100 g of
A.R. sodium chloride per dm^.

Reagent II. This consists of 2M-sodium hydroxide containing 20 g of A.R.
sodium chloride per dm^.

Zinc-ion solution. Dissolve about 7.0 g of A.R. zinc sulphate heptahydrate
in 25 cm^ of Reagent I.

Cadmium-ion solution. Dissolve about 6.0 g of A.R. crystallised cadmium
sulphate in 25 cm^ of Reagent I.

EDTA solution, O.OIM. See Section X, 50.

Buffer solution, pH = 10. Dissolve 7.0 g of A.R. ammonium chloride and
57 cm^ of concentrated ammonia solution (sp. gr. 0.88) in water and dilute to 100
cm^.

Solochrome Black indicator mixture. Triturate 0.20 g of the solid dyestuff
with 50 g of A.R. potassium chloride.

Xylenol Orange indicator. Triturate 0.20 g of the solid dyestuff with 50 g of
A.R. potassium chloride (or nitrate). This solid mixture is used because solutions
of Xylenol Orange are not very stable.

Nitric acid, ca. IM.

Procedure. Wash the anion exchange column with two 20-cm^ portions of
Reagent I; drain the solution to about 0.5 cm above the top of the resin. Mix
thoroughly equal volumes (2.00 cm^ each) of the zinc- and cadmium-ion
solutions and transfer by means of a pipette 2.00 cm^ of the mixed solution to the
top of the resin column. Allow the solution to drain to within about 0.5 cm of the
op of the resin and wash down the tube above the resin with a little of Reagent I.
ass 150 cm^ of Reagent II through the column at a flow rate of about 4 cm^ per
minute and collect the eluate (containing the zinc) in a 250-cm^ graduated flask;
1 ute to volume with water. Wash the resin with about 50 cm^ of water to remove
most of the sodium hydroxide solution. Now place a 250-cm^ graduated flask in
position as receiver and pass 150 cm^ of IM-nitric acid through the column at a
ra e of about 4 cm^ per minute; the cadmium will be eluted. Dilute the effluent to
/50cm3 with distilled water.

e resin may be regenerated by passing Reagent I through the column, and

185 ,

VII, 14 QUANTITATIVE INORGANIC ANALYSIS

Ciin then be used again for analysis of another Zn-Cd sample.

Aimlvscs. (a) Original zinc-ion solution. Dilute 2.00 cm^ (pipette) to 100
cm^ in a graduated flask. Pipette lO.Ocm* of the diluted solution into a 250-cm^
conical fiask, add cn. 90 cm^ of water, 2 cm^ of the bufTcr solution, and suflicient of
the Solochrome Black indicator mi.xture to impart a pronounced red colour to
the .solution. Titrate with standard O.OIAf-EDTA to a pure blue colour (sec
Section X, 61).

(b) Zinc-ion cluatc. Pipette 50.0 cm^ of the solution into :i 250-cm^ conical
flask, neutralise with hydrochloric acid, and dilute to about 100 cm-^ with water.
Add 2 ern^ of the buffer mixture, then a little Solochrome Black indicator powder,
and titrate with standard 0.01 .M-EDTA until the colour changes from red to pure
blue.

(r) Original cadmium-ion solution. Dilute 2.00 cm-’ (pipette) to 100 cm-’ in a
graduated flask. Pipette 10.0 enr’ of the diluted solution into a 250-cm^ conical
flask, add CO. 40 cm^ of water, followed by solid hevamincand a few milligrams of
Xylenol Orange indicator. If the pH is correct (5~6) the solution will have a
pronounced red colour (see Section X, 61). Titrate with standard 0.01 Af-F,DTA
until the colour changes from red to clear orangc-ycllow'.*

(rfl Cadmium-ion eluate. Pipette 50.0 cm' of the .solution into a conical fla.sk,
and partially neutralise (to pll 3--4) with aqueous .sodium hydroxide. Add solid
hcxamineltogitcn pH of5-61and a litilcXylenul Orange indicator. Titrate with
standtud 0.0! M-nDT.‘\ to a colour change from red to dear orange-yellow.

Some typical results are given below.

0.200 cm' of original Zn*’ * solution required 17.50 cm' of 0.01038, \f-EDTA
Weight of Zn'‘ per cm’ -- 17,50 5 0.01038 .y 65.38--.^ 59.35mg
50.0cm' ofZn-' cluatc r 17.45 cm' or0.0I03S.\f.r;DTA.

Zir ’ recovered - 5 a 17.45 :-:0.0I038 x 65.38 .r- 59.21 mg
0 200cm' oforiginalCd' ‘ solution required l'I.27cm' of0.dl03SAf-I;DTA.

. . Weight ofCd'‘ per cm’ ^ 5 x 19.27 y 0.01038 -x 112.4= H2.4mg
50.0 cm' of Cd-’’ eluate r 19 35 cm' ofn.OHDSAf-HDTA.

C’cf' ' recovered 5 >• 19 35 0.01038 x 1 12.4 =•• 1 IIS mg

VII. 14. DETERMINAIIO.N- OF FLI’ORIDF: W ITH THi: AID OF A
C.A'riON FXCHANOlvR. Thrnry. Soluble mclnliic fluorides may be
analysed by passing an aqueous Mshuion titrough a cation exchange column in a
polythene tube, collcclmg the liberated hydrofluoric acid in a polythene beaker,
and titrating it with standard sodium hydroxide solution.

The student may determine the fluoride content of sodium fluoride to gain
experience in the determination,

Procciiurc. OIrtain a polythene tube, 125 cm long and 12 mm internal
diameter, provided witli a nor/lc at the lower end. I'ill the tube above the nozzle
with short lengths of polytlicnc tubing |15 mmx2 mm) stacked vertically to
provide a support for the resin. Attach a short length of tliin-wallcd PVC tubing
to the jet outlet, and then attach a length of about 10 cm polythene tubing (6 mm
internal and 10 mm external diameter) to the latter. Attach a pinch-cock orscrew
clip to the thin-wallcd PVC tubing; this will enable the flow of liquid to he
stopped at will. Charge the column in the usual manner with Zcrolit 225,

TIic soluiion may also t)c tilraled at pU « 10 using SoUKhiornc Ill.ick as indicator.

186

ION EXCHANGE VII, 15

hydrogen form (volume about 15 cm^); leave the column full of water to just
above the bed'of resin. Prepare a ca. O.lM-sodium fluoride solution, using an
accurately weighed amount of the dry A.R. salt. Pass 25.0 cm^ of this solution
through the column followed by 4 x 15 cm^ of boiled-out distilled water (or de-
ionised water) and collect the effluent in a polythene beaker. Maintain a rate of
flow of about 4 cm^ per minute. Titrate the total effluent with standard O.lM-
sodium hydroxide, using phenol red or phenolphthalein as indicator.

Calculate the fluoride content of the sample of sodium fluoride.

Vn, 15. DETERMINATION OF SULPHUR IN IRON PYRITES WITH
THE AID OF A CATION EXCHANGER. Theory. The sample of iron
pyrites is dissolved in a mixture of concentrated nitric and hydrochloric acids.
After dilution (and filtration, if necessary), the solution is passed through a cation
exchanger (sulphonic acid type) in the hydrogen form. The effluent contains
hydrogen ion as the only cation. The sulphate is determined by precipitation as
barium sulphate. The barium sulphate may either be weighed or dissolved in
excess of standard EDTA solution and the excess titrated with standard
magnesium chloride solution (Section X, 75).

Reagents. Barium chloride solution, ca. 0.05M. Prepare from the A.R.
solid.

EDTA solution, 0.05M. See Section X, 50.

Magnesium chloride, 0.05M. Prepare from pure magnesium (Section X,

62).

Bujfer solution, pH = 10. Add 7.0 g of A.R. ammonium chloride to 57 cm^
of concentrated ammonia solution (sp. gr. 0.88) and dilute to 100 cm^ with water.

Solochrome Black indicator. See Section X, 28.

Procedure. Weigh out accurately about 0.50 g of iron pyrites* and treat it
with 10 cm^ of a mixture of 3 volumes of concentrated nitric acid and 1 volume of
concentrated hydrochloric acid in a 250-cm^ beaker. Allow the reaction to
proceed at room temperature for 30 minutes, then warm the covered beaker on a
steam bath until all reaction appears to cease, remove the clock-glass cover, and
evaporate the solution to dryness. Treat the residue with 5 cm^ of concentrated
hydrochloric acid and evaporate to dryness again. Dissolve the residue in 1-2
cm of warm concentrated hydrochloric acid, dilute to about 100 cm^ with hot
water, and filter through a sintered-glass crucible (G3). Wash the residue with hot
water, and combine the washings with the filtrate.

Percolate the combined solutions through a 25-cm column (contained in a
burette or tube with overflow as in Fig. VII, 3, a) of a cation exchange resin (e.g.,
erolit 225) in the hydrogen form; pass water through the column until the
effluent is neutral. Maintain a flow rate of about 3 cm^ per minute and collect the
6 uent in a 500-cm ^ graduated flask. Finally, dilute the solution to the mark with
water. Pipette 25.0 cm^ of the solution into a 250-cm^ beaker, dilute to 50 cm^,
eat to boiling, and add a slight excess of 0.05M-barium chloride (about 12 cm^)
Keep on a steam bath for 1 hour. Filter through a filter-paper disc

atman No. 542) supported on a Gooch porcelain crucible, and wash the
precipitate with cold water. Transfer the precipitate and filter paper back to the

<ie°ermina\lo British Chemical Standards) may be used for practice in this

187

\1I. 16 QUANTITATIVE INORGANIC ANALYSIS

original beaker. Introduce 35.0 cm*^ of standard 0.05A/-EDTA into the beaker,
followed by 5 cm^ of concentrated ammonia solution: boil gently until the
precipitate dissolves (about 10 minutes). Dilute thcclcar solution to 100 cm^, add
4 cm'^ of buficr solution and a few drops of Solochromc Black indicator. Titrate
the excess of EDTA with standard 0.05A/-magncsium chloride until the colour
changes from blue to wine red.

Calculate the percentage of sulphur in the .sample of iron pyrites.

Some typical results arc given below.

Weight of iron pyrites = 0.5001 g

EDTA .solution = 0.04697AI. MgClj .solution = 0.05160A/.

Volume of EDTA solution = 35.00 cm^

Mean titre of excess of EDTA = 26.75 cm^ of MgCL solution.

S present in 25.(X)cm^ of solution =

{(35.00 X 0.04967j -(26.75 x 0.05160)J x 32.06 = 0.3582 x 32.06 mg
Per cent of S in iron pyrites - (0.3582 x 32.06 x 20 x 100)/500.1 45.9j.

Thcanaly.scd Rid,sdalc sample contained 46.1 percents.

VIl, 16. SEPARATION OF COBALT AND URANIUM FROM MIXED
AQUEOUS-ORGANIC SOLVENT USING A CATION EXCHANGE
RIuSIN. Theory. The increased selectivity of ion exchange resins which may
be achieved by the use of mixed aqueous-organic solvent .systems is illustrated by
the separation of uranyl ion from cobaltfll) using a strong acid attion c.xchangc
resin. It has been shown (Ref. 6) that uranium, as UOj' ■* , in a mixture composed
of 90 per cent tetrahydrofuran and 10 per cent 6A/ nitric acid (v/v) has a much
lower distribution coefiicient than have most other di- and even higher-valent
ion.s. This forms the ba.sis of the separation of small amounts of cobalt from
relatively large amounts of uranium. The low distribution coedicient of UO-^ ‘
is probably due to the formation of an anionic niiraio complex UOilNOj).,"
which may be eluted from the cation exchange resin as an ion association
complex with tetrahydrofuran.

CII.-CII,

CH;-CU,

rCtl.-LIt;

^O-II
LCII.-Cit,

NO,,-

ruintr

(THI'Ii)*NOj +UO.(NO,)j rrnrH)‘tUO;(NOPj) '

Cobalt is retained on the resin and can subsequently be eluted with a mixture of
90 per cent tetrahydrofuran and 10 per cent 6Af-hydrochloric acid.

Reagents. Mixed solreni (.4). 90 per cent tetrahydrofuran 4- 10 per cent
OA-f-nitricacid. Prepare from pure reagents.

Mixed solvem (B), 90 per cent tetrahydrofuran -HO per cent 6M-
hydrochloric acid. Prepare from pure reagents.

Cation exchange rexin. Zerolit 225 (M ^ form).

Sample .solution. Dissolve 2 g uranyl nitrate hexahydrate and 5 mg
cobalt(Il) in 10cm-’ of mixed solvent (A).

Procedure. Equilibrate the resin {ca. 5 g) by allowing it to stand in the
mixed ,solvcnt(A), ca. 20 cm’, for about 30 minutes. Prepare a small column (5 cm
X 1 .0 cm) of the resin and introduce the sample solution on to the top of the resin
column. Elute uranium with the mixed solvcnt(A) at a flow rate of about 2 cm’

188

ION EXCHANGE VII, 17

min'* until the eluate is no longer yellow (50 cm^ of mixed solvent should be

sufficient). • ■

With the mixed solvent(B) {ca. 100 cm^) elute the cobalt using a similar flow
rate. Collect the blue eluate in a 100 cm^ graduated flask and make up to the mark
with more mixed solvent(B).

Measure the absorbance of the cobalt solution against the mixed solvent(B) at
675 nm. Calibrate the spectrophotometer using solutions of 2, 4, 6, 8, 10 mg of
cobalt(II) in 100 cm^ of mixed solvent(B).

Vn, 17. DETERMINATION OF URANIUM WITH THE AID OF A
LIQUID ANION EXCHANGER. Theory. The formation of an anionic
sulphate complex by uranium(VI) in relatively dilute sulphuric acid solution
provides the basis for separation of uranium from solutions containing high
concentrations of iron salts. Uranium is extracted from a sulphuric acid solution
using a chloroform solution of the liquid anion exchanger, Amberlite LA.1. Back
extraction of uranium with sodium carbonate solution gives an alkaline solution
which reacts with hydrogen peroxide to give yellow peruranate. This selective
reaction enables uranium (10-100 mg) to be determined spectrophotometrically
by measuring the absorbance of the solution at 410 nm (Ref. 13).

Reagents. Sodium carbonate solution, 100 g dm“^. Prepare from A.R.
solid and de-ionised water.

Hydrogen peroxide, 20 vols.

Liquid anion exchanger solution. Dissolve 4 cm ^ of Amberlite LA. 1 in pure
chloroform and make up to 100 cm® with this solvent.

Uranium solution. Prepare a standard uranium solution by dissolving
0.524 g A.R. uranyl nitrate hexahydrate, U 02 (N 03 ) 2 . 6 H 20 , in 250 cm® 0.5M~
sulphuric acid.

Procedure. Add increments of 10, 20, 30, 40 cm® of the standard uranium
solution to beakers each containing 90 cm® of sodium carbonate solution. Heat
each solution and boil for about 5 minutes, cool and dilute to 200 cm® in a
graduated flask. Transfer a 25 cm® aliquot in each case to a 50 cm® graduated
flask, add 5 cm® of hydrogen peroxide (20 vol.) and dilute the solution with water
to the mark. Measure the absorbances of the solutions at 410 nm using a 2 cm cell
(it is advisable to de-gas the solutions by shaking them thoroughly prior to
measuring their absorbances). Prepare a calibration graph of absorbance against
uranium concentration.

Prepare a sample solution containing approximately 25 mg of uranium(VI)
and 500 mg iron(III), (5 g ammonium iron(III) sulphate may conveniently be used
tor this second ion), in 50 cm® of IM sulphuric acid and dilute to 100 cm® with
water. Transfer the solution to a separating funnel (250 cm®), add 30 cm® of anion
exchanger solution and shake for 30 seconds. Allow the two phases to separate
and run the lower chloroform layer into a clean beaker. Repeat the extraction of
the aqueous phase with two further 30 cm® portions of anion exchanger solution
and coinbine the three chloroform extracts. Wash out the separating funnel and
h 'v ‘Combined extracts to it. Add 30 cm® of sodium carbonate solution and
Make for 30 seconds. Allow the two phases to separate, run the lower chloroform
ayer back into the original beaker and transfer the sodium carbonate extract to a
ean beaker. Return the chloroform solution to the funnel, add another 30 cm® of
ca h ™ solution and repeat the process. Make a third sodium

onate extraction and combine the sodium carbonate extracts.

189

Vn, 18 QUANTITATIVE INORGANIC ANALYSIS

Heat (lie combined solutions to boiling and boil for about 5 minutes, cool and
dilute to 200 cm^ in a graduated flask. Transfer a 25 cm^ aliquot (if necessary filter
the solution through a dry Whatman No. 541 paper) to a 50 cm^ graduated flask,
add 5cm^ of hydrogen peroxide (20 vol.)and dilute the solution with water to the
mark. Measure the absorbance of the solution against water at 410 nm using a 2
cm ceil. Compare the value obtained with the etdibration graph previously
prepared.

VII. 18. CONCENTRATION OF COPPER(II) IONS FROM A BRINE
SOLUTION USING A CHELATING ION EXCHANGE RESIN.
Theory. Conventional anion ;ind cation exchange resins appear to be of limited
use for concentrating trace metals from saline solutions such as .sea water. The
introduction of chelating resins, particularly those based on iminodiacetic acid,
makes it possible to concentrate trace metals from brine solutions and separate
them from the major components of the solution. Tlius the elements cadmium,
copper, cobalt, nickel and zinc are selectively retained by the resin Chclex-100
and ciin be recovered subsequently for determination by atomic absorption
spcctrophotomctr>'(Rcf. I4).TocnbancethesensitivityofthcAAS procedure the
eluatc is evaporated to drj’ness and the residue dissolved in 90 per cent aqueous
acetone.* The use of the chelating resin oflers the advantage over concentration
by solvent extraction that, in principle, there is no limit to the volume of sample
which can be used.

Reagents. Standard copperill) solutions. Dissolve 100 mg of spectro-
scopically pure copper metal in a slight excess of nitric acid and dilute to 1 dm^ in
a graduated flask with dc-ionised water. Pipette a 10 cm^ aliquot into a 100 cm^
graduated flask and make up to the mark with acetone (A.R.); the resultant solu-
tion contains 1 0 gg of copper per cm^. Use this stock solution to prepare a series
ofstandardsolutionscontaining 1. 0-5.0 /ig of coppcrpcrcm^, each .solution being
90 percent with respect to acetone.

Sample solution. Prepare a sample solution containing 100 //g of coppcrfll)
in 1 dra^ of O.SAf -sodium chloride solution in a graduated flask.

Ion exchange column. Prepare the Chelcx-100 resin (100-500 mesh) by
digesting it wath excess (about 2-3 bed-volumes) of 23f-nitric acid at room
temperature. Repeat this process twice and then transfer suflicient resin to fill a
1.0 cm diameter column to a depth of 8 cm. Wash the resin column with several
bed-volumes of de-ionised water.

Procedure. Allow the whole of the sample solution (1 dm^) to (low through
the resin column at a rate not exceeding 5 cm^ min' *. Wash the column with 250
cm'^ of de-ionised water and reject the washings. Elute the coppeijll) ions ss'ith 30
cm-* of2M-nitricacid, place the eluatc in a small conical flask (100 cm^, preferably
silica) and evaporate carefully to dryness on a hot plate (use a low temperature
setting). Dissolve the residue in 1 cm^ ofO.lM-nitric acid introduced by pipette
and then add 9 cm^ of acetone. Determine copper in the resulting solution using
an atomic absorption spectrophotometer which has been calibrated using the
standard copper(II) solutions.

Note. All glass and silica apparatus to be used should be allowed to stand

In the illustrative experiment <Jc.scril>ed here, coppcr(ll) ions in a brine solution are concentrated
fromO.I Pp.m. to about .tjp.p.m. prior lodclerminationbyatomicabsorptionspcctrophotometry.

190

ION EXCHANGE VH, 19/20

overnight filled with a 1 : 1 mixture of concentrated nitric and sulphuric acids and

then thoroughly rinsed with de-ionised water. This treatment effectively removes

traces of metal ions. . ; '

VII, 19. References. , ,,

1. C. B. Amphlett (1964). Inorganic Ion Exchangers. Amsterdam; Elsevier.

2. J. Inczedy (1966). Analytical Applications of Ion Exchangers. Pergamon Press; 1st
English edition.

3. L. R. Snyder (1965). Chromatog. Rev., 7, 1.

4. G. J. Moody and J. D. R. Thomas (1968). Analyst, 93, 557.

5. W. R. Heumann (1971). Ton exchange in non-aqueous and mixed media’, CRC crit.
Rev. analyt. Chem., 2, 425.

6. J. Korkisch (1966). ‘Combined ion exchange-solvent extraction (CISE): A novel
separation technique for inorganic ions’, Separation Science, 1 (2, 3), 159.

7. H. P. Gregor et al. (1952). Ind. Eng. Client., 44, 2834.

8. E. Blasius and B. Brozio (1967). ‘Chelating Ion-exchange Resins’, in (ed. H. A.
Flashka and A. J. Barnard), Chelates in Analytical Chemistry, Vol. 1, p. 49. New
York; Marcel Dekker.

9. G. Schmuckler (1965). ‘Chelating Resins — Their Analytical Properties and
Applications’, Talanta, 12,281.

10. C. F. Coleman et al. (1962). Talanta. 9, 297.

11. H. Green (1964). ‘Recent Uses of Liquid Ion Exchangers in Inorganic Analysis’,
Talanta, 11, 1561.

12. H. Green (1973). ‘Use of Liquid Ion-exchangers in Inorganic Analysis’, Talanta, 20,
139.

13. H. Green (1964). ‘Determination of Uranium in Cast Iron’, BCIRA Journal, 12, 632.

14. J. P. Riley and D. Taylor (1968). ‘Chelating Resins for the Concentration of Trace
Elements from Sea Water and Their Analytical Use in Conjunction with Atomic
Absorption Spectrophotometry’, Anal. Chim. Acta, 40, 479.

Vn, 20. Selected bibliography

1. 0. Samuelson (1962). Ion Exchanger Separations in Analytical Chemistry. New York;
John Wiley.

2. G. H. Osborn (1961). Synthetic Ion-Exchangers. 2nd edn. London; Chapman and
Hall.

3. E. Lederer and M. Lederer (1957). Chromatography. Division II. ‘Ion Exchange
Chromatography.’ Amsterdam; Elsevier.

4. L. Meites and H. C. Thomas (1958). Advanced Analytical Chemistry. Ch. II. Ion
Exchange and Chromatographic Methods. New York ; McGraw-Hill Book Co.

5. R. Kunin (1958). Ion Exchange Resins. 2nd edn. New York; John Wiley.

6. J. E. Salmon and D. K. Hale (1959). Ion Exchange. A Laboratory Manual. London;
Butterworths.

7. G. W. Ewing (1960). Instrumental Methods of Chemical Analysis. Ch. 19. Ion
Exchange. 2nd edn. New York; McGraw-Hill Book Co.

8. Ion Exchange Resins (1971). 5th Edition, 3rd Impression (revised). Poole, Dorset; The
British Drug Houses Ltd.

9. W. Rieman and R. Sargent (1961). Ion Exchange, in W. G. Berl. Physical Methods in
Chemical Analysis. Vol. 4. New York ; Academic Press.

• R. Kunin, F. X. McGarvey, and A. Farren (1956). ‘Ion Exchange’, Analytical
Chemistry, 28, 729.

11. R Kunin, F. X. McGarvey, and D. Zobian (1958). ‘Ion Exchange’, Analytical
Chemistry, 2% ()%\.

12. R, Kunin (I960). ‘Ion Exchange’, Analytical Chemistry, 32, 67R.

191

VII, 20 QUANTITATIVE INORGANIC ANALYSIS

13. J. A. Marinskyand Y. Marcus (cils.) (1973). ‘Ion Exchange and Solvent Extraction —
A Series of Advances’, New York; Marcel Dekker.

14. F. Hclffcrich (1962). ‘Ion Exchange', New York; McGraw-Hill.

15. M. Qureshi rto/. (1972). ‘Recent Progres.sin Ion-exchange Studies on Insoluble Salts
ofPolybasic Metals’, Separation Science,!, 615.

16. W. Riemnnnand H. F. Walton (1970). Ion Exchange in Analytical Chemistry. Oxford;
Pcrganion Press.

17. J. Inc7edv(1972). ‘Use ofion exchangers in analytical Chemistr>’, Rev. anaivt. Chem.
1,157.

192

PAPER, THIN LAYER AND
CHAPTER VIII COLUNIN CHROMATOGRAPHY

Vin, 1. GENERAL INTRODUCTION. Chromatography has been defined
as primarily a separation process which is used for the separation of essentially
molecular mixtures. It depends upon the redistribution of the molecules of the
mixture between two or more phases. The various types of chromatography
include adsorption chromatography, fluid partition chromatography, and ion
exchange. The main systems employed in jijrtitipp .chxomatogfaphy are; gas
partition (see Chapter IX), liquid partition employing fixed beds (i.e. column
chromatography), thin layer and paper chromatography. In each case distri-
bution takes place between a ‘stationary’ sorbed ‘liquid’ phase and a mobile fluid
in intimate contact with it. In Jiquid partiti ^udHuagiatography, a mobile liquid
phase flows over an essentially stationary liquid phase sorbed on a support; in
paper chromatography the support is paper or treated paper, whereas in thin
layer chromatography the adsorbent is coated on a glass plate or plastic foil. We
shall deal only with selected aspects of partition chromatography upon cellulose
with particular reference to inorganic analysis.

The simplified technique for paper chromatography will be evident from the
following description of a classical experiment designed to separate a mixture of
amino acids. A strip of Whatman No. 1 filter paper, about 25-30 cm long and
1.5 cm wide, is marked lightly with a pencil line about 5 cm from one end. The
mixture, containing 5-15 micrograms (^g) each of glycine, alanine, valine, and
leucine in 2-4 microlitres (^1) of total solution, is spotted from a capillary pipette
on to a marked spot A in the middle of the pencil line (Fig. VIII, la). The solvent
is allowed to evaporate. The developer is prepared by shaking together liquified
phenol and water in about equal quantities in a separatory funnel for several
minutes ; the phases are allowed to separate cleanly, and the upper aqueous layer
is drawn off into the dish, which is placed in the bottom of the large gas jar (Fig.
VIII, 1 fe— not drawn to scale). The paper is hung in the gas jar with the upper end
held in the glass trough. The liquid is introduced to saturate the air in the gas jar
with water and phenol in the ratio that will be in equilibrium with the developer;
after a while the paper becomes conditioned to the reagents. (This conditioning
of the paper is not usually necessary in inorganic chromatography.) the
developer, which is the lower phenolic phase, is now introduced into the glass
trough and the gas jar is closed. The developer moves by capillary action into the
paper and, aided by gravity, passes down over the mixture at A, and development
proceeds. The bulk mobile phase is phenol containing water, and the thin
stationary phase is water (containing some phenol) sorbed on the paper. After

• J

193

Mil, 1 QUANTITATIVE INORGANIC ANALYSIS

(a) (b) fri

FiR. vin, 1

(lie front of (lie developer has moved to almost the lower edpc of the paper, the
piis jar is opened, the paper removed, and the position of the front B marked
immediately (Fig. VIM. Ir). The paper is then allowed to dry. Tlie amino acids
are colourless and. in order to reveal the position of the ?oncs. a colour reaction
with ninhydrin is used. The paper is sprayed with a 0.1 per cent .solution of
ninhydrin in buttmol and wtirmed fora short time to accelerate the reaction with
the amino acids. The amino ticid ?ones appear as red-purple spits (Fig. VIll. It/).
The resulting chromatogram is described and the zones are characterised by
values. The Rf value is rlcfined by the relation :

^ _ distance (cm) fro^n smrting line to centre of zonj^

'' ~ distance (cm) from starling line to solvent front

The R, value measures the velocity of movement of the zone relative to that of the
developer front. The measurement is made by measuring the distancc.s from the
starting line (centre of initial mi.xed zone) to the developer front and the centre of
density of each zone; thus for zone 1. R, ~ (Fig. VTIl. Ir/). The R^ values
will identify the amino acids, and the intensity of the zone may be used as a
measure of the concentration by comparison with .standard spots.

Chromatography on cellulose is basically a solvent-extraction type process;
the materials to be separated undergo partition between the aqueous phase held
in the inert cellulose malric and the organic solvent used as the mobile phase.
Those components of the mi-xturc to be .separated which are most readily soluble,
in the organic mobile phase will have /?, values near, or equal to, unity. Tlio.se
components which have a lower solubility in the organic phase will have /If
values near to zero. The Ry value is characteristic of a particular species in any
given type of separation, and is sometimes used for the qualitative identification
of the unknowTt species. In simple cellulose chromatography the mechanism is
largely partition in type; adsorption processes play only a small part (c.g., ducto
the presence of small quantities of carboxyl groups in the paper), but this cflcct is
not normally apparent when strongly acidic solvents are employed.

In inorganic scparalions on unmodified cellulose two main groups of factors
appear to govern the mobility of different elements. First, those factors which
increase their solubility in the organic phase and thus lead to higii Ry values.

lOd

PAPER, THIN LAYER AND COLUMN CHROMATOGRAPHY VIII, 1

Inorganic salts are not usually soluble in organic solvents and such solubility is
often indicative of complex formation. With solvents containing donor oxygen
atoms in the presence of a small amount of hydrochloric acid, those metals which
form chloro complexes move readily, probably owing to the solubility of the free
chloro-complex acid in the organic solvent. Thus iron(III) is readily mobile with
solvents containing hydrochloric acid while nickel(II) does not move appreciably .
this behaviour may be compared with the different absorption of these metals on
to anion exchange resins from hydrochloric acid solution. The second group of
factors governing mobility are those which tend to cause low Rp values, i.e.,
retain the components of the mixture in the stationary aqueous phase of the
cellulose. This behaviour is apparent with metals if anions are present in the
mixture which form strong water-soluble complexes or which give an insoluble
precipitate. Such interference may often be overcome by prior elimination of the
offending anion (by precipitation) or by the addition of an otherwise inert species
which complexes the interfering anion more strongly than do the components of
the mixture to be separated. .

Reference should be made to modified cellulose obtained by the introduction of
diethylaminoethyl groups, of carboxyl groups and of phosphate groups into the
cellulose matrix. Such substituted celluloses possess The ^ advantages of
normal ion exchange materials with the additional merit that they can be
obtained in sheet form, thus enabling standard paper-strip methods to be. used.
Cellulose phosphate appears to be an excellent selective exchanger'; thus thorium
is taken up by cellulose phosphate from 4W-acid, and' so rendering possible the
recovery of this element from sulphuric acid solutions of monazite. Iron(III) ions
and uranyl ions behave similarly, and can be eluted . only by the use of a
complexing agent such as ammonium carbonate.

Cellulose phosphate is essentially cellulose dihydrogen phosphate, and is a
bifunctional exchanger containing both strong acid (thus simulating strongly
acidic cation exchange resins) and very weak acid groups. Carboxymethyl-
cellulose may be regarded as a weakly acidic cation exchanger which functions
most readily at a pH above 4-5. Diethylaminoethyl-cellulose resembles in base
strength the corresponding tertiary-amino ion exchange resins; it does not
function in strongly alkaline solutions. As in most other cellulose derivatives, the
majority of the substituted groups are located accessibly close to the surface of
the exchanger, thus facilitating the exchange with large molecules more readily
than an orthodox resin exchanger. The modified cellulose ion exchangers are'
marketed* in both paper sheet and in powder form. ' . .

There are three main methods of conducting inorganic chromatographic
separations on cellulose, viz., by the use of (a) glass plates or plastic foils coated
with thin layers of cellulose, (b) paper strips, and (c) columns of cellulose. Thin
layer and paper strip methods are essentially micro-analytical in character and
cannot usually be employed with more than 100 fig of sample. For quantitative
work the plate or strip may be sprayed with a reagent which forms coloured
complexes with the elements present ; the quantities of these elements can then be
estimated by visual comparison of the spots produced with those obtained using
standard solutions of the elements under identical conditions. Alternatively the
portions of the cellulose thin layer,, or paper , strip, . containing the spots may be

* By the manufacturers of Whatman filter papers, W. & R. Balston and Co Ltd. ;

195

Vni, 2 QUANTITATIVE INORGANIC ANALYSIS

suitably treated to extract the separated elements fordetennination by spcctro-
photomctric or other appropriate instrumental methods.

The cellulose column method is generally employed for macro-scale work, it
being usual to condition the column by prior equilibration with the solvent to be
used for the separation. The most convenient procedure is to use conditions
under which the element to be determined is eluted first from the column; it may
then be determined by a standard method, either titrimctric or instrumental.

It will be appropriate at this point, following the above general introduction to
the topic, to enlarge upon the two particular aspects of inorganic thin layer
chromatography and of the development of high performance liquid chromato-
graphy. ^

Vni, 2. THIN LAVER CHROMATOGRAPHY. The tcch giami-of-tlua
''^layer chromatography (TLC) use s an adsorbent coa ted on a glass plntc^sjhe_
staiionars' "pTiase Tihdnj c^opntcnt of TiTc cTTromaiogfam takes place as the
"mo bile p lmsc percolates t'firbugh The adsorbcntTTfiTrnaycrchrom’atogr.iplTy has,
well knovsTirdL^iitcVaTlvaiTfagcs ovcT^iperchromatography because of its
U convenience and rapidity, its greater sharpness of separation and its Jhigh
ZU ZschsTtivity . A bnenicsc'riprion oi cxperhncmal procedure wlTBc given herewith
particular reference to the separation of cations.

Preparation of the plate. In thin layer chromatography a variety of coating
material s arc available, althoug h silica gel is used more often than other materials.
The .separation of cations on silica gel is not. however, always 'satisfactory' as
many cations have similar Rf values and remain grouped together on this
adsorbent. Cellulose powder is recommended as an adsorbent for separation of
cations by TLC even though separations may be slower than those obtained on
silica gel. The use of cellulose powder may be regarded as a substitute for paper
chromatography and data obtained for inorganic paper chromatography are
generally applicable to inorganic TLC on cclluiosc.

Tiiin la yers of cellulose can be ma de ln:_sprca ding a n atmcous-slurry of
celluFo^ppw^^using one of the commercially available applicators.* It is most
■ Tfnportant that thc^lass plaTcs to wnich tiuTThirTniycnsTo be applied shoiild be
tlmrpughly clean andTT[iBlCh'irTfcI2i'ccum|ni^vc O~T3y~wa ^^ iTiFplateT^in a
cbncenlratecf sbVlium carbonate solution followed by tlioro^T nhslng wTh
"distilled water.

"nrhc aqueous slurry of cellulose powder is prepared by mixing about 15 g
powder in 90cm-’ of distilled water and dispering the powder for about 1 min
using a mechanical mixer. I’hc cellulose powder used for inorganic TLC is of a
special microcrystallinc nature.! In partition chromatography, unactivated
plates arc used and cclluiosc layers thus require no activation by heating. The
coated plate may be dried overnight at room temperature.

Rcady-io-usc thin layers, prepared with the most widely used adsorbents, arc
now available, c.g. as prccoatcd glass plates and plastic foils. Plastic sheets
precoated with cellulose (which may also incorporate fluorescent material) are

* Available from Sbandon Scienlific Company Ltd, Camlab (GbKS) Lid, and from GrifTm and
George Ltd.

T Supplied by E. Merck Laboratory Chemicals (distributed in the UK by .■Xndemian and Co Ltd),

»56

PAPER, THIN LAYER AND COLUMN CHROMATOGRAPHY VIII, 2

marketed* and are very convenient for inorganic TLC work as they can be cut to

the required size. ' , ,, * •

Sample application. The sample solution to, be applied, should contain
between 0.1 and 10 mg of the cation per cm^ and may be neutral or dilute acid;
about 1 n\ of solution is applied with a microsyringe or micropipette near one end
of the chromatoplate (about 1 .5-2.0 cm from the edge of the plate) and the latter
air dried. Equilibration of the chromatoplates is not.necessary and development
of the plate can start immediately after it is dried.

Development of plates. The chromatogram is usually developed by the
ascending technique in which the plate is immersed in the developing solvent
(redistilled or chromatographic grade solvent should be used) to a depth of
0.5 cm. The tank or chamber used is preferably lined. with-sheet^-offilter-paper
which dip into the solvent in the base of the ch a mber ; this ensures that the
chamber is saturated with solvent vapwr (Fig. VIII, 2). Development is allowed
to proceedTintil'the'soh^t front has travelled the required distance (usually
10-15 cm), the plate is then removed from the chamber and the solvent front
immediately marked with a pencil line.

Plate in the—
course of
development

Solvent

Paper round sides of tank

Fig. VIH, 2 Reproduced from D. Abbott and R. S. Andrews (1965). An Introduction to
C/iromarogra/p/iy. London; Longman.

The positions of the separated solutes can be located by various methods.
Coloured substances can be seen directly when viewed against the stationary
phase while colourless species may usually be detected by spraying the plate with
an appropriate reagent which produces coloured areas in the regions which they
occupy. Some compounds fluoresce in ultraviolet light and may be located in this
way. Alternatively if fluorescing material is incorporated in the adsorbent the
solute can be observed as a dark spot on a fluorescent background when viewed
under ultraviolet light. (When locating zones by this method the eyes should be
protected by wearing special protective goggles or spectacles.) The spots located
by this method can be delineated by marking with a needle.

Thin layer chromatography using microscope slides. The separation of
cations may be very conveniently achieved using cellulose-coated microscope
slides (or precoated cellulose sheets cut into pieces about the size of microscope
slides). Very small spots (diameter ca. 1 mm) of sample solution are introduced
onto the plate, e.g., by using a pointed paper strip saturated with the solution.

* Manufactured by the Eastman Kodak Co.

197

Vni,3 QUANTITATIVE INORGANIC ANALYSIS

The plate is air dried and developed in a small bottle until the solvent front has
travelled 2-3 cm. Separation is achieved in 5-10 minutes and the separated
cations may be located by the usual procedures.

Quantitative inorganic thin layer cliromatograpliy. Quantitative analysis of
separated constituents on thin layer plates is generally carried out by measure-
ment of the photodensity and area of the spot, i.c. by photodensitomcio' of the
plate.* This type of procedure requires comparison with spots obtained using
known amounts of standard mixtures which must be chromatographically
examined on the same plate as the sample.

An alternative procedure involves removal of the scptiratcd components from
the plate by scraping off the relevant portion of the adsorbent. The component is
eluted (extracted) from the adsorbent with a suitable solvent and determined by
an appropriate physical technique, e.g., by ultniviolet. visible, or fluorescence
spectrophotometry.

3. HIGH PliRFORMANCK LIQUID CHROMATOGRAPHY. No
ly account of contemporary chromatographic techniques would be complete
without mention of high performance liquid chromatography (HPLC). Classical
column liquid chromatography is a well established separation procedure in
which the mobile liquid phase flow.s slowly through the column by means of
gravity. The method is generally characterised by Iqv^olttmn-cnicicncics and
long se paration tim es. .Since about 1969, however, ther^ha.s been a very marked
ren^ of interest iifi the technique of liquid column chromatography due to the
development by Kirkland and Huber of high pressure systems operating at
pressures up to 2.07 x 10'' Nm"’ (3000p.s.i.) (Ref. 1). In this method small
diameter columns ( 1 -3 mm) with support particle sizes in the region of 30 /mi arc
used and tlic eluent is pumped through the column at :i high (low rate (cm. 1 to
5cm^min'‘). Separations by this method may be effected much more rapidly
(about 100 times faster) than by the usc pfconycntionaI.!iquid chromatography.

Although the currently available cbmmercia[cquipmcntt is rather expensive,
HPLC has already shown itself to have wide applicaHonTiTbrcanic chemistry.
The development of inorganic applications is likely, e.g., in the field of ion
exchange chromatography where pellicular rc.vins. that is, resins produced as thin
coatings on the surface of spherical glass beads (20-50 micrometres diameter),
are now available commercially under the trade name 'Zipax', The heads have a
porous surface about 2 micrometres thick which serves to bond the resin coating.

The tipplication of high performance ion exchange chromatography for the
separation of inorganic compounds is illustrated by the separation of iransplu-
tonium clcracms; the method is based on sequential elution of the cations from a
cation exchange column with an anionic coniplexing agent. Exposure of the rc.sin
to radiation from the radioisotopes is reduced since a satisfactory separation is
achieved in a much shorter lime, and raiiiation induced damage to the resin is
minimised (Ref 2).

A low co'ii dcn<.itomclcr using fibre optics has been developed bv Kontes, In.strumcnt Group,
bpruce Street. Vineland, New Jersey, USA. 08360.

t R^sve Angel Ltd now market an economical familiarisation kit for 1 il’LC vvhicb can be operated up
to 2.4 X 10 Nm (350 p.s.i.). Store advanced equipment is supplied bv Waters Assocs.. Perkin
Elmer Ltd, etc.

198

PAPER, THIN LAYER AND COLUMN CHROMATOGRAPHY VIII, 4

Vm, 4. SEPARATION OF NICKEL, MANGANESE, COBALT AND ZINC
AND DETERMINATION OF' Rp VALUES. Discussion. This experiment
illustrates the separation of the above four metals using either thin layer or paper
strip techniques. A sraalf measured sample of the solution (e.g. from an Agla
micrometer syringe' or a micropipette) is placed iiear one end of the thin layer
plate or the paper strip and the chromatogram developed using a mixed
acetone-hydrochloric acid solvent. After the solvent front has moved a suitable
distance the plate or paper strip is removed from the tank or jar, the solvent
evaporated after marking the solvent front and the metal ions rendered visible in
the form of coloured spots or .bands by spraying with a suitable reagent. The
experiment permits the evaluation of Rf values which are approximately Ni 0. 1 ,
Mn 0.25, Co 0.55, Zn 0.9. For quantitative work the bahds can be cut from the
paper strip, the ihetals extracted and determined ; . alternatively arid rhore
conveniently a reflectance densitometer may be used (see footnote to page 198).
The procedures used for quantitative analysis by thin layer chromatography
have already been described.

Reagents. Nickel-, manganese-, cobalt-, and zinc-ion solution: Prepare this
from the following A.R. salts: (NH4)2S04,NiS04,6H20; MnS04,4H20;
CoCl2,6H20 or (NH4)2S04,CoS04,6H20; and ZnS04, 7H2O. Dissolve approp-
riate quantities in 2A/-hydrochloric acid to give solutions containing 1 0 iig each
of Ni^'*', Co^'*', and Zn^'*' in 0.01 cm^ of solution.

Acetone-HCl solvent. Mix 43.5 cm^ of A.R. acetone, 4cm^ of concentrated
hydrochloric acid (sp. gr. 1.18), and 2.5 cm^ of water.

Spraying reagent PACFjR (trisodium pentacyanoammine-ferrat'e/rubeanic
acid). Dissolve 0.70 g of the pentacyanoammine-ferrate in 20 cm^ of water and
pour the resulting solution into a solution of 0.25 g of rubeanic acid in 1 0 cih* of
ethanol. Shake the mixture for 1 5 minutes and filter. The filtered solution is ready
for use. The reagent should be prepared on the day it is to be used.

Preparation of Na2[Fe(CJ^iNH:i]6H20. , Weigh out lOg of finely-ground,
sodium nitroprusside (disodium pentacyanonitrosyl ferrate) into a sihall conical
flask, and add 24 cm^ of concentrated ammonia solution, sp. gf; 0.88. Shake well
and loosen the cork so that gas can escape; when all the solid has dissolved and
gas evolution has begun, place the flask in a refrigerator at -7°C for 48 hours.,
Warm to room temperature, add a little moire concentrated ammonia solution,
sp. gr. 0.88, filter at the pump, remove as much liquid as possible, and then wash
the precipitate with a small volume of methanol.' Rapidly remove as much as
possible of the methanol by suction, transfer the product to a glass dish in a
desiccator over calcium chloride, and keep it in the dark. The yield is 5.4 g.' '
Acetic acid, ca. 0.2M. Dilute 1 1 cm^ of glacial acetic acid to 1 litre with water.

Procedure A. Separation using paper strips! Pour 25cm^ of the
acetone-HCl solvent into the tall gas jar or 2-litre measuring cylinder (Fig; VIII,
1) and an equal volume into the upper solvent container.* Allow to stand for at
least 15 minutes before commencing the separation. Fold a 30-cm strip of
Whatman’s No. 1 paper (2.5 cm wide) about 3 cm from one end. Draw a thin
pencil line about 2.5 cm from the fold and mark the mid point. Pipette 0.05 cm^ of

► The external dimensions of the Pyrex boat are about 40 x 25 x 35 mm (internal capacity cn. 30 cm’)
wrtiM^rodTs 6mm

199

vm, 5 QUANTITATIVE INORGANIC ANALYSIS

the test solution on the strip; hold the paper vertically and draw the tip of the
micro-pipette (or syringe) along the pencil line and near the mid point. Prepare a
second strip in the same way and allow both to dry in the air for about 30
minutes. Hang the slrip.s in the jar with the folded ends in the upper solvent
container, and allow the solvent to diffuse within 2.5 cm of the bottom of the
paper strips (about 3 hours). Remove the strips, allow the solvent to evaporate,
and e.spose to ammonia vapour to neutralise the acid. Spniy with PACF/R
reagent ; use an atomiser. Remove the excess of the reagent by dipping llic paper
strips into dilute accticacid (0.2^/). The metals arc visible ascolotired bands: Ni,
blue; Mn, blue: Co, brown: and Zn, red.

Determine the R,- values as dc.scribcd in Section VIII, 1.

Procedure P. Separation by thin layer chromatography. Prepare a cel-
lulose coated glass plate, or more conveniently cut out a strip (5 x 20cm) from a
plastic sheet precoated with cellulose. Draw a thin pencil line about 2cm from
one end of the .strip (or plate), apply a 1 0 /d sample of the metal ion solution to the
centre of the line using a microsyringe and allow to air dry. Place the strip upright
in the developing tank so that the sample end is immersed in the developing
solvent to a depth of 0.5 cm; the solvent should be introduced into the lank at
least 15 minutes before commencing the scpanition. Allow the development to
proceed until the solvent front has travelled about IScm (this takes approxi-
mately 30 minutes), remove the strip (or plate) and allow about H) minutes for
the solvent to evaporate. Neutralise excess acid by exposing the strip to ammonia
vapour for about 10 minutes and spray with pentacyanoammineferratc/ru beanie
acid reagent using an atomiser. Remove excess rciigent as described under
Procedure A ; the metals .arc visible as similarly coloured spots and the/?,- values
can be determined.

VIH, 5. SEPARATION OF NICKEU COPPER. COHALT, AND ZINC.
This provides an alternative experiment to that of the previous Section.

Reagents. XickeP. copper-, eobttih. ami zine-ioii soltiiioii. Prepare this
from the following A.R. salts: (NH4),S0j.NiS04.6H20; CuS04.5Hj0;
CoClj.dHjO or (NH4),S04CoS04,()IKd; and ZnSOj.flDO. Appropriate
quantities arc dissolved in distilled water to give a solution containing about 1 mg
of each metal ion in I cm^ of solution.

.Solvent. Prepare a mixture of acetone, ethyl acetate, and 6Af-hydrochloric acid
in the ratio of 9:9:2 by volume from the A.R. reagents.

Sprayinp reopen!. Prepare a O.I percent solution of rubcanic acid in ethanol.

Procedure. Apply lO/d of the solution of metal ions to either a paper strip
or cellulose coated plate or foil. Allow to air dry and continue the separation as
described under cither Procedure A or B (previous .section). After development,
remove the paper strip or tlic plate (foil), allow the solvent to eraporate, and
expose it to ammonia vapour to neutralise the acid. Spray the paper strip (both
sides) or thin layer plate (foil) with the rubcanic acid reagent. Nickel is rendered
visible as a bluc-purpic band, cobalt as a yellow-orange band, and copper as an
olive-green band. Tlie zinc is visible as a pink band by spraying with a dilute
solution ofdithiz.onc in chlorofomi.

Evaluate the Rj. values of the four ions by the method described in Section

’ ni^ 1 •

200

PAPER, THIN LAYER AND COLUMN CHROMATOGRAPHY VIII, 6

Vni, 6. SEMI-QUANTITATIVE SEPARATION OF COPPER, COBALT,
AND NICKEL ON SLOTTED PAPER STRIPS. Discussion. A speciaUy cut
rectangular sheet (21.3cmx 11cm) of Whatman’s fflter paper, provided with
eleven slots (3 mm x 9 cm) cut into the paper parallel to the short side so as to
leave twelve strips 1.5 cm wide joined at the top and bottom (Fig. VIII, 3), is

employed’^. This permits the separation of a
number of samples simultaneously with very simple
apparatus and is especially valuable for semi-
quantitative trace metal analysis; the technique is
useful for geochemical prospecting. The metals are
detected by spraying the strips with suitable
reagents, and the amounts present are detected by
visual comparison with standards. Thus copper,
cobalt, and nickel may be determined after a single
separation. .

An aliquot of the solution is applied to one end of each strip of paper so that,
by leaving the two end strips vacant, ten sample solutions may be placed on the
sheet. The volume of test solution used is about 0.01 cm^, arid it is applied with'
the aid of a capillary pipette so that it spreads right across the strip to form a thin
rectangular patch. The sheet is bent so that it forms a cylinder and clipped at the
upper end with a paper clip, preferably of polythene. After suitable drying, the
sheet is placed vertically in a covered beaker containing the solvent, the depth of
which must not exceed 1 cm. The solvent is allowed to diffuse up the strip and
reaches the level of the top of the slots in 1 0-35 minutes, the time depending upon
the solvent. The sheet is removed just before the solvent reaches this level, and the
positions of the metals on the strip are located by various treatments of the sheet,
according to the metals to be determined. For the chromatographic separation of
copper, cobalt, and nickel from each other and from other elements, a mixture of
pure ethyl methyl ketone (75 parts), concentrated hydrochloric acid (15 parts),
and water (10 parts, v/v) is satisfactory. With this solvent, the approximate
values are : copper, 0.65 ; cobalt, 0.45 ; and nickel, 0.10.

Reagents. Solvent. Mix 37.5 cm^ of pure ethyl methyl ketone, 5 cm^ of
water and 7.5 cm^ of concentrated hydrochloric acid (sp. gr. 1 .1 8).

Copper solution. Dissolve 1.12g of A.R. copper(II) chloride dihydrate in
50 cm^ of concentrated hydrochloric acid (sp. gr. 1 . 1 8) and 5 cm^ of concentrated
nitric acid (sp. gr. 1 .42), and dilute to 1 00 cm^ with water. -

Cobalt solution. Dissolve 1.70 g of A.R. hydrated cobalt chloride in 50 cm^ of
concentrated hydrochloric acid and 5 cm^ of concentrated nitric acid, and dilute'
to 100 cm^ with water.

Nickel solution. Dissolve 1.86 g of A.R. nickel sulphate hexahydrate in 50 cm^
of concentrated hydrochloric acid and 5cm^ of concentrated nitric acid and
dilute to 1 00 cm^ with water.

All the above solutions contain about 0.42 g of each metal in lOOcm^ of
solution. Prepare a standard mixed Cu, Co, Ni solution containing ca. 0 1 g of
each metal in 1 00 cm^ by placing 25.0 cm^ of each of the above solutions in a 1 00-
cm graduated flask and diluting to the mark with a mixed acid solution (50 cm^
of concentrated hydrochloric acid 5 cm^ of concentrated nitric acid -f 45 cm^

* This is marketed as Whatman Pattern F-SF 2925 CRL.

Fig. Vin, 3

201

Vin, 7 QUANTITATIVE INORGANIC ANALYSIS

of water). Use this .standartl solution to prepare a series of solutions as follows;

Solution No. I 2 3 4 S 6 7 S 9 10

Sl:indartl Cu. Co. Ni solution, cm^ 10 9 8 7 6 5 4 .1 7.5 4.5

Mixed acid solution, cm^ 0 I 2 3 4 5 6 7 2.5 5.5

In order to gain c.xpericncc in the dctcnninalion. solutions I to 8 will be
regarded as standards, and .solutions 9 and iO a.s unknowns.

Ruhcaiiic add solution. Dissolve 0. 1 g of rubcanic acid in 60 cm^ of ethanol and
dilute to 100cm’ with water.

Procedure. Pour 25 cm’ of tlie ketone .solvent into a 600-cm’ Pyrex beaker
and cover it with a Petri dish. Py means ofa graduated micropipcife add 0,01 cm’
of each of the solutions No.s. i to 10 evenly along lines about 2cm from one side
of the .sheet of slotted paper (Whatman C.R.L. pattern): each paper contains
twelve strips, but do not use the two outer .strip.s, I'orm the slicct into a cylinder
and secure the two upper ends together with a paperclip. Place the cylinder with
the sample spots lowermost in a clean, drx' beaker in a boiling water bath : remove
the papercylindcr after 3 minutes and stand it immediately in the solvent beaker,
replacing the Petri dish cover. .Allow the .solvent to dilTusc to the top of the strips
(time: up to 60 minutes), remove the cylinder from tiic beaker, and allow it to dry
in the air for 5 minutes. Place the papercylindcr inside a. second 600-cm’ Ircakcr.
fitted with a Petri dish cover and containing a small {30cm’) beaker charged with
concentrated ammonia solution ; leave for 3 minutes. Take the cylinder from the
beaker, lay the .sheet Il.it on plate glass, and .spray* it lightly on both sides with the
rubcanic acid solution. Allow the sheet to dry in the air. Compare (visually) the
copper (grey-green), cobalt (orange), and nickel (blue) bands for solutions 9 and
10 with tho.se produced by the standards, and in this way c.stimatc their Cu. Co,
and Ni content.s.

Vin,7. SEPARATION OFIRON AND ALUMIXIUM ON A CELLULOSE
COLUMN. Theory. The ions of irondll) and of aluminium may be sep-
arated on a cellulose column using A.R, acetone containing 2 per cent of
hydrochloric acid as solvent : the irondll) forms a chloro complc.x and is readily
mobile in the sol vent, probably as the metal chioro-complex acid. The aluminium
remaining in thecellulosecolumn may then bcclutcd with 1 .)(-hydroch!oricacid.
The acetone -hydrochloric acid solvent may be used to .separate a number of
binary mi.xture.s : Ni, Pb. Al. Cr, Ti, Zr, and Th (immobile) from Co. Cu. Cd. Fc,
Zn, Mn. V, and Mo (mobile).

Rcagenl.s. IronfUlJ .saltuion. Dissolve sufiicient ,A.R. ammonium
iron(ll!) sulphate, accurately weighed, in S,)/-hydrochloric acid to give a
solution containing about 25 mg of irondll) per cnr'.

Ahmunium-ion solution. Dissolve sufiicient A.R. aluminium ammonium
sulphate, accurately weighed, in SAZ-hydrochloric acid to give a solution
containing about 25 mg of aluminium per cm’.

Standard EDTA solution, ca. 0.02M. See Section X, 5ft.

Standard :inc sulphate solution, ca. 0.02M. Dissolve pure zinc in dilute
sulphuric acid.

Variaininc Blue B indicator. Prepare a I per cent aqueous solution.

An cxccitcni spray is murfcctcd by Shnndon Scientific Co Utl, bv Baird and Tatlock (London) Lid,
and by Microcticmical Spccialilics Co of Berkeley. 10, Catifoniia.

202

PAPER, THIN LAYER AND COLUMN CHROMATOGRAPHY VIH, 7

- Solochrome Black indicator. See Section X, 28. .

Water repellent for glassware. A 2 per cent solution of dichlorodimethylsilane
in carbon tetrachloride is satisfactory.

Organic extraction solvent. Add 6cm^ of concentrated hydrochloric acid to
300 cm^ of A.R. acetone. (The latter usually contains up to about 0.7 per cent of

water.) . . . , ■

Whatman’s 'ashless tablets'. For preparation of cellulose pulp.

Preparation of cellulose column. Prepare cellulose pulp* by boiling Whatman
‘ashless tablets’ with an aqueous solution containing 5 cm^ of concentrated nitric
acid (sp. gr. 1.42) per 100 cm^ for 2-3 minutes; decant, remove the excess of acid
by washing with water. Then wash with ethanol, followed by diethyl ether.

Preparation of glass extraction tube. Use a Pyrex tube,, 30-40 cm long, about
16 mm internal diameter, and drawn out at the lower end to an internal diameter
of about 3 mm. It is necessary to treat the inside surface of the glass extraction
tube with a water-repellent material to prevent ‘wall’ effects due to creep of the
aqueous test solution at the top of the column down between the glass tube and
the cellulose. Shake a small amount of the water-repellent solution in the clean
tube until the whole of the surface has been treated, remove the excess after a few
minutes, and wash the tube with ethanol, followed by a little water. In. the
absence of fluoride, the ‘silicone’-treated tube retains its water-repellent
properties for a large number of separations. ,

Procedure. Place a large glass bead {ca. 6 mm diameter) at the bottom of
the glass column (to act as a support for the cellulose) ; fit the narrow end of the
tube with a short length of rubber tubing, a screw clipi and a short glass tube.f
Fill the column to about two-thirds of its length with the organic extraction
solvent. Introduce some cellulose pulp or powder. ,With the aid of a glass rod,
which has one end flattened to form a plunger of slightly smaller diameter than
that of the glass column, beat the cellulose up to form a smooth slurry. Allow the
solvent to run out until most of the cellulose has settled ; add more solvent and
cellulose pulp (or powder) and beat the cellulose up with the glass plunger.
Repeat the process until a ‘settled’ column of about 12 cm length results. Press
the cellulose gently down until the rate of flow of solvent through the column
decreases to about 2 cm^ per second. Maintain the solvent level above that of the
‘settled’ cellulose. Finally, run 50 cm^ of the solvent through the column to
remove any traces of iron present in the cellulose: close the screw clip when the
level of the liquid is just above that of the cellulose column.

Mix equal volumes (say, 5.00 cm^) of the prepared iron and aluminium
solutions and pipette 2.00 cm^ of the resulting solution into a 100-cm^ beaker.
Add sufficient dry cellulose pulp (or powder) to form a friable mass when mixed,
and transfer the wad to the top of the column; rinse the beaker with 5 cm^ of the
organic solvent and transfer to the column. Gently beat the wad of cellulose with
the glass plunger and pack it down to form a continuous part of the column. Rest
the glass plunger in the sample beaker. Detach the screw clip : collect the eluate in
a 350-cm^ conical flask. Introduce 150cm^ of the solvent into the column in 5-
cm^ portions ; use each portion to rinse the beaker which contained the sample

Add 50 cm^ of water to the eluate ; this contains all the iron.

Whatman Cellulose Powder, Standard Grade, may be used directly, and is more convenient >
t Alternatively, a glass stopcock may be sealed to the narrow end of the glass column.

Vni, 7 QUANTITATIVE INORGANIC ANALYSIS

Replace the receiver by another 350-cin^ conical flask. Elute the aluminium
from the column with 100cm’ of 1 A/-hydrochloric acid, added in 5-cm’
portions. Remove the acetone from both cluaics by evaporation on a water bath.

Analysis ofeluates. (a) Iron solution. Add a few cm’ of 20-volumc hydrogen
peroxide to oxidise any ironfll) present to iron(IIl); boil for 10-15 minutes to
destroy the excess of hydrogen peroxide. Cool, dilute to 100cm’. add dilute
ammonia solution until” the pH is 2-3 (use Congo Red paper or a pH meter),
followed by 5-6 drops of Variaminc Blue B indicator. Titrate with standard
EDTA solution until the colour changes from blue-violet to yellow,

(b) Aluminium solution. Dilute thccluatc to 250 cm’ in a graduated flask. To
50.0 cm’ of this solution add 25.0 cm’ of standard EDTA solution, adjust the pH
to between 7 and 8 (use a pH meter or phenol red paper), and introduce a few
drops of Solochrome Black indicator. Titrate rapiiily with .standard zinc sulphate
solution until the colour changes from blue to wine red.

Siandanlisaiion of ironfUl) solution with EOTA. Dilute 2.00cm’ of the
iron(in) solution to lOOcm’ with water, add dilute ammonia .solution to the first
perceptible colour change of Congo Red paper (pH 2-3), followed by 5 drops of
Variaminc Blue B indicator. Titrate with standard EDTA solution until the
colour changes from blue-violet to yellow; the colour is grey just before the end
point.

Standardisation of aluminium-ion solution with EDTA. Dilute 2.00 cm’ of the
aluminium-ion solution to 100cm’ in a gradu.atcd flask. To 10.00cm’ of the
diluted solution add 25.0cm’ of standard EDTA solution, followed by lAf-
ammonia solution to a pH of 7-S. Introduce a few drops of Solochrome Black
indicator and titrate rapidly with standard zinc sulphate solution to the blue to
wine-red end point. It may be ncccssarj- to add a drop of ammonia solution to
maintain the pH above 7, otherwise the blue colour tends to fade.

Calculate the weight of iron(lll) and of aluminium ion in the volume of
solution employed.

Some typical results arc given below.

2.00 cm’ of the iron(in) solution required 40.70cm’ of0.01964,A/-EDTA;

2.00cm’ ofsolution contains 40.70x0.01964 x 55.85 = 44.67 mg Fe.

In aluminium-ion titration, excess of EDTA required 17.65cm’ of

0.01920A/-ZnS04.

.'. 2.00cm’ of the aluminium-ion solution contains

lOx ((25.00 x 0.01964) -(17.65 X 0.01920)} X 26.98 == 41.01 mgAl.

Iron recovered from 2.00 cm’ of mixed solution = 20.25 cm’ of0.01964.U-

EDTA = 20.25 x 0.01964 x 55.85 = 22.13 mg.

Aluminium recovered from 2.00cm’ of mixed solution: 25.00cm’ of

0.0I964A/-EDTA added; excess of EDTA = 17.70cm’ of 0.01920A/-

ZhSO^;

.'. Eluate contains

5x{(25.00 x 0.01964)-(17.70 x 0.01920))x26.98 = 20.40 mg At.

Alternatively the metal ions may be determined by atomic absorption
spectrophotometry. Suitable solutions arc obtained by collecting each eluate
(after removal of acetone) in a 1 dm’ graduated flask and diluting to the mark with
de-ionised water. The instrument should be calibrated with standard Fe’"^ and
A1 solutions covering the 0-50 p.p.m. concentration range.

204

PAPER, THIN LAYER AND COLUMN CHROMATOGRAPHY VIII, 8

Vm, 8. SEPARATION OF COBALT AND NICKEL ON A CELLULOSE
COLUMN. Theory. See Section Vm, 7. , .

Reagents. Cobalt-ion solution. Dissolve about 2.0 g of A.R. hydrated
cobalt chloride in 25 cm^ of 8 Af-hydrochloric acid.

Nickel-ion solution. Dissolve about 11.0 g of A.R. nickel sulphate hepta-
hydrate in 25 cm^ of SM-hydrochloric acid.

Standard EDTA solution, c&.0.02M.SQcSec\.{on\,5Q.

Xylenol Orange indicator. Prepare a 0.5 per cent aqueous solution.

Bromopyrogallol Red indicator. Prepare a 0.05 per cent solution in ethanol.

Buffer solution. Mix equal volumes of 1 M-anunonium chloride and 1 M-
aqueous ammonia solutions.

Organic extraction solvent. Add 6 cm^ of concentrated hydrochloric acid to
300 cm^ of A.R. acetone.

Procedure. Prepare a cellulose column about 15 cm long as detailed in
Section Vm, 7, use Whatman Cellulose Powder, Standard Grade.

Mix exactly equal volumes (say, 5,00 cm^ each) of the cobalt; arid nickel-ion
solutions. Pipette 2.00 cm^ of the mixed solution into a 100-cm^ beaker. Add
sufficient dry cellulose powder to the solution to form a friable mass and transfer
it to the top of the column as completely as possible ; rinse the beaker with 5 cm^
of the organic extraction solvent and add to the column. Gently beat the wad of
cellulose with the glass plunger, and pack it down to form a continuous part of
the column. Introduce 125 cra^ of the extraction solvent into the column in 5-cm^
portions ; use each portion first to rinse the beaker which contained the sample.
Collect the eluate in a 350-cm^ conical flask; the eluate contains the cobalt as the
deep blue chloro complex.

Now elute the nickel from the cellulose by passing 1 00 cm^ of 1 M-hydrochloric
acid through the column; add the acid in 5-cm^ portions. Collect the eluate in a
250-cm^ conical flask; this now contains the nickel.

Analysis of eluales. {a) Cobalt solution. Add 50 cm^ of water to the eluate and
remove the acetone by volatilisation on a steam bath. Boil the solution for several
minutes, cool and dilute to about lOOcm^ with water. Adjust the pH to 6 by the
addition of solid hexamine, add a few drops of Xylenol Orange indicator, and
titrate with standard 0.02M-EDTA until the colour changes from red to orange
yellow (compare Section X, 61).

(b) Nickel solution. Heat the eluate on a steam bath to remove acetone, cool
and dilute to 250 cm^ in a graduated flask. -Transfer 50.0 cm^ of the diluted
solution to a 250-cm^ conical flask, dilute to about lOOcm^ with water, nearly
neutralise with A.R. sodium hydroxide, add lOcm^ of the buffer solution, 10
drops of Bromopyrogallol Red indicator, and titrate, with standard 0.02Af-
EDTA until the colour changes from blue to red (compare Section X, 58). ■

Standardisation of cobalt solution with EDTA. Dilute 2.00 cm^ of the prepared
cobalt solution to 100 cm^ with water, adjust the pH to 6 by the addition of solid
hexamine, add a few drops of Xylenol Orange indicator, and titrate with the
standard EDTA solution until the colour changes from red to orange yellow.

Standardisation of the nickel solution with EDTA. Dilute 2.00 cm^ of the
prepared nickel solution to 100 cm^ with water in a graduated flask. Pipette
1 0.0 cm^ of the dUuted solution into a conical flask, dilute to 1 00 cm^ add 1 0 cm^
of the buffer solution, 10 drops of Bromopyrogallol Red indicator. Titrate with
standard 0.02M-EDTA to a colour change from blue to red

Calculate the percentage of cobalt and nickel recovered in the separation and

205

VIU, 9 QUANTITATIVE INORGANCC ANALYSIS

also the weights oftlicsc elements in the volume of sohilion employed.

Some typical results arc given below.

2.00 cm’ of cobalt-ion solution required 31.85 cm’ of 0.01975Af-I:DTA;

2.00 cm’ of solution contained 31.85 x 0.01957 x 58.94 = 36.73 mg Co.
0.200cm’ of nickel-ion solution required 15.03cm’ orO.01957 Af-EDTA;

2.00 cm’ of .solution contained 10 x 15.03x0.01957 x 58.69 - 172.63 mg Ni.
Cobalt recovered from 2.00 cm’ ofthemi.xcd solution s 1 5.95 cm’ of 0.0! 957 Af-
EDTA:

Weight of cobalt in 2.00cm’ of the mixed solution - 15.95 x 0.01957 x
58.94 " 1 8.39 mg

Nickel recovered from 2.00/5cm’ of the mixed solution s 14.95cm’ of
0.0 1957 A/-EDTA;

Weight of nickel in 2.00 cm’ of the mi.xed solution = 5 x 14.95
X 0.01957 X 58.69 85.86 mg.

VIU, 9. SEPARATION OF COPPER AND NICKEL ON A CELLULOSE
COLUM.N. Theory. See Section 3TU, 7.

Reagents. Copper-ion soiutinn. Dissolve about 2.8 g of A.R. coppcr(ll)
sulphate pentahydrate in 50 cm’ of8A/-hydrochloric acid.

Nickcl-ion .'iolulion. Dissolve 1 1 .0 g of A.R. hydrated nickel sulphate in 25 cm’
8A/-hydrochloric add.

Standard TDTA solution, ca. 0.02M. See Section X, 50.

Tast.sidphon Black /'(C.!. No. 26990) tW/rn/nr. Prepare a 0.5 percent aqueous
solution.

Bromapyroyallol Red indicator. Prepare a 0.05 per cent .solution in ethanol.
Organic extraction .solvent. Add 6.0 cm’ of concenlnitcd hydrochloric acid to
300 cm’ of A. R. accionc.

Procedure. Prepare a cellulose cohiinn about 15 cm long as described in
Section VIIL7; use Whatman Cellulose Powder, Standard Gnide,

Mix c.xactly equal volumes (say, 5.00 cm’ each) of the copper- and nickel-ions
solutions. Use 2.00 cm’ of the mixed .solution for the .separation exactly ns
detailed in Section VNI, 8. Elute the copper from the cellulose column as an
orange brown solution with 125 cm’ of the organic extraction .solvent, and the
nickel with 100 cm’ of 1 A/-hydrochloric aci<l.

.Analysis ofcluatc.s. (a) Copper solution. Add 50 cm’ of water to the eluatc,
ncutrali.se the solution with A.R. sodium hydroxide, and remove the acetone by
evaporation on a steam bath. Render the solution faintly acid ssath dilute
hydrochloric acid and boil for a few minutes. Cool the greenish-yellow solution,
add 50 cm’ of water, followed by 1 0 cm’ of concentrated ammonia solution and
10 drops of Fast Sulphon Black F indicator. Titrate with standard 0.02A/-EDTA
until the colour changes from blue-violet to bright green (compare Section X.56).
(b) Nickel solution. Proceed as in section VUI, 8.

Standardisation of the copper .solution with EDTA. Pipette 2.00 cm’ of the
copper solution into a 250-cm’ conical flask, dilute to 100 cm’ with water, and
just ncutr.alisc wth A.R. sodium hydroxide. Add 10 cm’ of concentrated
ammonia solution and 10 drops of Fast Sulphon Black F indicator. Titrate with
standard 0.02A/-EDTA to a colour change from blue-violet to bright green.
Standardisation of the nickel. solution with EDTA. See Section VDI, 8.
Calculate the percentage of copper and nickel recovered in the separation and
also the weights of these elements in the volume of solution employed.

PAPER, THIN LAYER AND COLUMN CHROMATOGRAPHY VIII, 10/11
Some typical results are given belowi

2.00 cm^ of the copper-ion required 29.97 cm^ of 0.01971M-EDTA;

2i00cm^ of solution contained 29.97 x 0.01971 x 63.54 = 37.53 mg Cu.

0.200 cm^ of nickel-ion solution required 15.03 cm^, of 0.01971M-EDTA;

.•. 2.00 cm^ of the solution contained 10 x 15.03 x 0.01971 x 58.69 = 173.86 mg
Ni.

Copper recovered from 2.00 cm^ of mixed solution = 14.90 cm^ of 0.01971Af-

Weight ofcopper in 2.00 cm^ of the mixed solution = 14.90x0.01971 x63.54
= 18.66 mg.

Nickel recovered from 2.00/5 cm^ of the mixed solution = 14.95 cm of
0.01971M-EDTA.

Weight of nickel in 2.00 cm^ of the mixed solution = 5 x 14.95 x 0.01971
X 58.69 = 86.47 mg.

Vin, 10. References

1. J. J. Kirkland (ed.) (1971). Modern Practice of Liquid Chromatography. New York;
Wiley-Interscience.

2. D. O. Campbell and S. R. Buxton ( 1 970). Ind. Eng. Chem. Process Design Develop., 9,
89; 9, 95.

Vin, 11. Selected bibliography

1. W. G. Berl (1951). Physical Methods of Analysis. Vol. II. Chromatographic Analysis.
New York: Academic Press.

2. C. R. N. Strouts, J. H. Gilfillan, and H. N. Wilson (1955). Analytical Chemistry, The
Working Tools. Vol. II. Ch. 27. Chromatography. Oxford : Clarendon Press.

3. F. H. Pollard and J. F. W. McOmie (1953). Chromatographic Methods of Inorganic
Analysis. London; Butterworths.

4. H. G. Cassidy (1957). Fundamentals of Chromatography, in A. Weissberger.
Technique of Organic Chemistry. Vol. 10. New York; Interscience.

5. E. Lederer and M. Lederer (1957). Chromatography. Amsterdam ; Elsevier Publishing
Co.

6. G. W. Ewing (1968). Instrumental Methods of Chemical Analysis. Ch, 18. Chroma-
tography. 3rd edn. New York; McGraw-Hill.

7. R. C. Brimley and F. C. Barrett (1953). Practical Chromatography. New York;
Reinhold.

8. I. M. Hais and K. Macek (eds.) (1964). Paper Chromatography. 3rd edn. New York;
Academic Press.

9. J. G. Kirchner (1967). Thin-layer Chromatography. New York; Interscience.

10. K. Randerath (1968). Thin Layer Chromatography. 2nd edn. New York; Academic
Press.

11. E. Stahl (ed.) (1964). Thin Layer Chromatography— A Laboratory Handbook. New
York; Academic Press.

12. F. W. H. M. Merkus (1970). ‘Progress in Inorganic Thin-Layer Chromatography’, in
Progress in Separation and Purification, ed. E. S. Perry and C. J. Van Oss Vol 3
p. 233, New York; Wiley.

13. F. H. Pollard, K. W. C. Burton and D. Lyons (1964). ‘Thin-layer Chromatography in
Inorganic Chemistry’, La6. Procn'ce, 13, 505.

14. L P. Garel (1965). ‘Thin-layer Chromatography III. Application in Inorganic
Chemistry’, Bull. Soc. chim. France, 1899.

15. R. C. Denney (1976). A Dictionary of Chromatography. London; MacmUlan.

207

Vin, II QUANTITATIVE INORGANIC ANALYSIS

16. J. Michal (1974). Inorf;miir Chronwiograplik Aiiti/vsh. London; Van Nosirand
Reinhold.

17. L. R. Snyder and J. J. KirUand (1973). Introihiciinn to Modem Liquid Chroma-
to^raphy. Chichester, Sus'.e.x; John Wiley and Sons.

18. J, C. Toiichsonc (cd.) (1973). Quaniiiaiire Thiti-Uiyer Chromatoyrapby. Chichester,
Su.ssc.x : John Wiley and Sons.

208

CHAPTER IX MS CHROMATOGRAPHY

IX, 1. INTRODUCTION.

mixtur e is separated into its_

stationary sorbent, ji tie technique is thus similar to liquid-liquid chfomato-
graphy except that mobile liquid phase is-veplaced by a moviryg gas phase. G as -

chromatography isMivided into, two major categories: gas-liquid chromato-
graph v (GLCl. where separation occurs by partitioning.a sample between a mobiie3 ::i=^
"gas phase and a thin la«er of non-volatile liquid coated on an inert support,
gas-sofid chromato'^^phy (GSC), which employs a solid ojf large surface areSM
the stationary phase. The present chapter deals with gas-liquid chromatography
and'some of its applications in the field of inorganic analysis, particularly in the
gas chromatography of metal chelates. Before considering these applications,
however, it is appropriate to describe briefly the apparatus used in, and some of
the basic principles of, gas chromatography. For more detailed accounts of these
topics the texts listed in the bibliography at the end of this chapter should be
consulted.

^Gas chromatography is a process by which a
(Constituents by a moving gas phase passing over a

IX, 2. APPARATUS. A gas chromatograph {see block diagram Fig. IX,
l(fl)} consists essentially of the following parts :

Flowmeter

Gas

cylinder

Heated

GLC

column

►-Sample

Chart I
recorder!

Pe aks correspond to ^ *N,

r'mbiridtral' components" detected J

. i

LI 1

Detector

Time-

Fig. IX, 1 (a) Block diagram of a gas chromatograph, (b) Typical chart record

Reproduced by permission from R. C. Denney (1970). The Truth About Breath
Tests. London; Nelson.

(1) a supply of carrier gas from a high-pressure cylinder. The carrier eas
used is either helium, nitrogen, hydrogen or a rgon, the choice of
(Jnjactors such as availabi lity, purity required/wnsumption and the type of

209

^■ . '^ASfcHROMATOGIlAPHY IX,2

gro 1 jfF~^^Eea to a large non-polar sKefeton, e .g., esters of high molecular
weight alcohols such as dinonyl phthalate.

3 . Polar compounds containing a relatively large proportion of polar groups ,
e.g. th e carbowaxesTpolyglycoIsTT

4. Hydro^rPSondihg class, i.e . polar liquid phases such a s glycol, glycerol^
bvHrhxvaHiffe-RT E :. Which possess an appreciable ^ number of hwrog giL- atQITlS .
_5^able tor hydrogen bonding.

Studies with meFal chelates have shown the best results to be obtained with the
silicones, e.g., SE-30 silicone gum rubber. .

The column packing is prepared by adding the correct amount of liquid phase —

dis^olvedlnaluitableTolvent (e.g., 'acetone or dichloromethane) to a weighed

quantity of the solid support in a suitable dis h. The volatile solvent is removed
either by spontaneous evaporation or careful heating, the mixt u re being gently ,
agitated to ensure a uml'orm distnbutioli of the liquid phase in the support . Final
' traces of the solvent may be removed under vacuum and the column packing re-
sieved to remove any fines produced during the preparation. The relative amount
of stationary liquid phase in the column packing is usually expressed on the basis
of the per cent by weight of liquid phase present, e.g., 15 . per cent loading
indicates that 100 g column packing contains 1 5 g of liquid phase oh 85 g* of inert ,
support . The solid sunoort should remain free flowing after being coated with the
liquid pGase! _ , -

(4) the'detector. The function of the detector, which is situated at the exit of
the separation column, is to sense alnd measure the , small amounts of the
separated components present in the carrier-gas-stream-leaving.the colum n. The
butput t rom the detector is~ted to a recorde r which produces a pen-trace callSd a
' Chromatogram ( r ^ig. IX. 1 The choice of detector will depend on factors such

aTThe concentra^iK level to be measured and the nature of the separated
components. Thi^ for work in the microgram ra nge the detector most generally
, used is probably me thermal conaucuviry cell, i^le f or, work ranging dow n fd'
the picogram le vel le.g.. ultra-trace analysts of metals}^™®!? sensitive detectors
such as those b^ed on~'ion'isation phenomena a re required^ liFTletectuis-nTO^
widely used iiT the gas chrdtnatb^aphj^jdFTnetal ■cKelSms are the thenhal
conductivity, flame ionisation and electron capture detectors; aiid a brief

idescriptipn-ofthese will be given. ' ...

O Ji^J FhCTmal conductivity detector. Thermal conductivity cells of katharo-
meters are th e most widely "used detectors in . gas chromatography. These
detectors employ a heated m etal fila ment or a thermistor (a semiconductor of
fused metal oxides ) to sense changes in the thermal conductivity of the carrier ga s
stream. Helium andhydrogen are the best carrier gases to useV conjunction witlT
this typ^tldetectorfsince their thermal conductivities are much higher than any

other gases A)n safety gtounds ^lium is preferrcfl hpransp nt

In the defector two pairs oShatched filaments are arranged, m^a^Whe^tone
bridge circuit ; two filaments in opposite arms of the bridge are su rrounded by the
car rier gas only, .while the othertwo ntaments are surrounded hv ths p.ttinpnt , (
_from the chromatographic column. Th is type of thermal pnnHv^tivit7~ceins
lUustrated in Fig. IX, 2(a) with two gas channels t hrough the cell: a sample
channgLaBl a_ reference channel. When pure ca rrier gas passes ov^^bTuTTiTi^ '
reference an d sample_filament.s7thp lSridge ir~RalanceH. -huTw hen

emerges from the column the rate of cooling^^e sample filam^rs'chan'^ ^^

211

IX, 2 QUANTITATIVE INORGANIC ANALYSIS

detector employed. ThuVi^liuin is preferred when thermal conductivity
detectors a }c,cniptoyed,beco.iisejol,nOij^hjH ern i Bcondti ciiyitv tcH
j hr vnpniiiN of mo st organic compounds. A ssociated with this higli pressure
supply of carrier gas arc thc^ijtcndant pressure regulators and flow meters to
control and monitor t he earner gas flow: flic operating;" elTiacney~t)TTRr'
apparatus is very^Icpcmlcnt on the maintenance of a constant flow of carrier gas.

(2) sample in}c ction system. Li qtiitl samples a rcJuilQilllcct L using a
microsvnnge wit h liyp.Qslertni c ncc dl^iic latter is inserted thr ough a self- sealing
silicone rubber septum and the sa mple injcclctl smoothly im Qji.hcat ed~m Hal
bl ock at tlieliSui'drth cT'.oJuiniL.Manipulation of the syringe may be regarda) as
an art developed with practice and the aim must be to introduce the sample in a
reproducible manner. The Icinpcrat urc of the sam plcp orLsI lould he suc h that the
liquid is rapidly vaporised b ut wi thout cither dLxmmposingT'Frntc iipnaling^^
sample; a uscTu! uitc or TIuTmb" uTTo 'set ' the sample panf temperature,
^ ap proxininlclv to the boiling Point of the least volatile compon entyFQryreaiBi'
^Iliciency, the sm:nT «r posMlmT sainple ji?^ 1 to 1 0 /7l ) consi stent with detector
scnsiihaty shouIcTlsc iisecT "

' (3)TiiFcOTinm'.I7l he actual separation of .•iamplc components is efi'cctcd in
the column where the nature of the solid support, type and amount of liquid
pliasc, metitod of packing, lengilt and temperature arc important factors in
obtaining the desired re.solution. Analytical columns arc usually prepared with 2
to 6 mm internal diameter glass tubing or 3 to 1 0 mm outer diameter metal tubing
which is normally coiled for compactness. Glass columns must be used if any of
the sample components .ire decomposed by contact with metal.

The ni;i tcriul ..c luisen ;is die inert su p port should be of uniform g ranular sire
and have good haiu ilinii.ehar act cnst!C5~ti..c. fie .strong enough not to break down ‘
m handling) and be capable of be ing packed into a nnifonn bed in a column. The
surface area of tlie tnatenal s hould be large so as to promote d i s t ri bu lion ^ the ’
jiquid phase as a li lm aniTcnsirrc' firc'r.TpiJ iltTainmcn'i of eqiulibrmm between
iit^onary iiodjmrTulcj^iscsTTT e mpsL ^huhonly usetl .sup ports (c.g. Cciitc) are
jiKuIiTTroirniimomaccous materia ls whi clTcan hold liquid phases in amounts
exceeding 20 per cent without becoming too sticky lo flow freely and can bccasily
packed.

Commercial preparations of these supports arc available in narrow mesh-
range fractions ; to obtain particlc.s of uniform si 7 e the material should be sieved
to the desired particle si/e range and repeatedly water floated to remove fine
particles which contribute to excessive pressure drop in the final column. To a
good approximation the height equivalent to a theoretical plate (see Ref. 12) is
proportional to the average particle diameter so that theoretically the smallest
possible particles should be preferred in terms of column cfiicicncy . Decreasing
jniij| cle si/e will, however, rapidl y increase the necessary appl jcdjinsjitc.giUDiy
ac hiev’c flow tHrou gluhcxaljLuntLi md in p ractice thiTb est choice is 80/100 mesh.

^3mm Q. column. It may be notctl hcrc tli^' TiTr'HIcHivc^^ of an y
-—^ISinj lic internal d iametcr-QfjJie^ibi ng shoulcTl i c at least eight timcs~Tfic
dmmeter of the solid support particles.

TTf^lcctiptnjrThrinbsTs^^^ liquid phase for a particular separation is
cnicufl. Liquid phases can be broadly chussified as follows :

.hyclroc arhon-tvnc liquid phases, c.g., paraffin oil (Nujol).
squalane,.Agi£;yy^Jgrcase an(V.rilicpn£j>um rubbcT; jlicTaittcr is used fo r hig h

210

jq-zTTr^ evvtarue / Ko

' 'aritv which possess a polar bi^polansable

2. _C_omp5unds of intermediat e polaril ^ _

grotIT!i att^Ee? to a large non-polar si^eton, e .g., esters of high molecular
weight alcohols such as di nonyl phthalate.

3 . Polar compounds c ontai ning a r elatively large proportion of polar groups
e.g. the carbowaxes

4. H^^re^ipSonding class, polar liquid phases such a s glycol, glycerol,
which D0ssess~ ah appreciable _ number of nvdrown atoms!

Available for hvHrogen bonding.

Studies with metal chelates have shown the best results to be obtained with the
silicones, e.g., SE-30 silicone gum rubber.

The column packing is prepared by adding the correct amount of liquid phase _
dissolved inTsuitable solvent (e.g., acetone or dichlorometEa ne) to a w eighed
quantity of the solid' support in a suitable dis h. The volatile solvent is removed
either by spontaneous evaporation or careful heating, the mixtur e b e in g, gently
agitated to ensure a umtorm distribution of the liquid phase in the support . Final
traces of the solvent may be removed under vacuum and the column packing re-
sieved to remove any fines produced during the preparation. The relative amount
of stationary liquid phase in the column packing is usually expressed on the basis
of the per cent by weight of liquid phase present, e.g., 15 per. cent loading
indicates that 1 00 g column packing contains 1 5 g of liquid phase bn 85 g of inert
support . The solid support should remain free flowing after beiiig coated with the
liquid pfiaseT^m ^ ^ •’

(4) the^eteSor, The function of the detector, which is situated at the exit of
the separation column, is to sense and measure the smalf amounts of the'
separated components present in the carger-gas-streamJeaving the colum n. THe'
' outpu t trom the detector is ted to a recorder which produces a pen-trace caned a
Ch romatogram (b ig. IX, l(b) ). .'l'hechoiceo^f detector will depend on factors such
asTlTe concentra^a. level to be measured and the nature of the separated
components. Thi^for work in the microgram range the detec tor most generally
used is probably tne thermal conauctmty ceilT^^ile f or work ranging down to
the picogram le vel (e.g.. ultra-trace analysts of metalsJ Trooje s ensitiye detectors
such as those based, on ionisation phenomena are requifed?!;! lieTlutectui s'-iiiust

widely used ifiTthe gas cnrolnatogfaphjfJ^t^meta'l 'cheldtes are the thermal
conductivity, flame ionisation and electron capture detectors, and ’ a brief

_^scriptmB-of these will be given. ' '

j )v_J HtCTmaI conductivity detector. Thermal conductivity cells or katharb-
meters are th e mpst widelv~used detectors in gas chromatography. These
detectors employ a heated metal filament or a thermistor (a semiconductor of
fused metal oxides ) to sense changes ifi the thermal conductivity of the carrier ga s
stream, helium and' hydrogen are the best carrier gases to useV conjunction wltlT

~this typ^fidetectortsince their thermal conductivities are much higher than any

other gases fen safety gynunds^^^im is preferrpH hpna use nt it<;

In the defector two pairs ofmatched laments are arranged in' a 'Whe^tone
bridge circuit ; two filaments in opp osite arms of the bridge are surrounded by the
carrier gas only-, .while the other Pro niaments are'surrounded by the effluent /
Jrg m the chromatographic column. Th is type of thermal cnnH^ictivity~ceirig
ill^ustrated in Fig. IX, 2(fl) with two gas channels through the cell: a s ample
channgLand a, reference channel. When pure carrier gas
jefe ^ce and sam pleJk m e nt . slhTBn d ge Ts halan ceg7T3m~^;4^ 5,

PfY>prcrf»c f'ViA _ 1 ' — — ‘

emerges from the column the rate of cooling of^e sample filammts chanji^jS^ /

IX, 2 QUANTITATIVE INORGANIC ANALYSIS

- Carrier , • (
jas yj^

sample ’

I I Leads

Delcclor

ouipul

Jr'X .

H>drorcn name
IXdarircd jcl
Insulalors
Ilody (earthed)

tr 1 1) df open

} Carrier pas + sample

■cJ. , / V^A -ir-'*'" ICarrier pas + sample

c-;\ > / w

Mr. I.\, 2\i4 Ti.cnrt.al conduciitlly dctinrlor (^flnmc lonlTnllon defeclor

corrosionrfr^Icins which”? jj illiticsjjjay, however, pose

..-X^?tivii^,hc n,ctnl halidS ^'Pnihag^HWcai chelates, owing to the

which M^iffid^TnTTTTr^V'i:^.':^'^^^^ produce a f lame

pnicnTTiTc 7 [,f. l»w ionisation

cin'rcTirTncasur^ofc .

usiiiiillyir^c or crirTT^iJ^iri^ — P^^^t:&!i^^£=£.lPl*i^2<]£J^McJJleJnodeJs•,:-

diagrammatically in^Fi^TD^Tlp^ /

that ♦’w produces ve r^Lfejy ions so L

sation occurs and ihcro i« -i i., * p^i'^ontaining compounds :irc present ioni-
namc. Because the samnic is in the electrical conductixaty of the

employed when furtWT^c.xanunat'iS'inhl’'' f'*"’'"-' s'rcam-spliiting device is
inserted between thecohirim necessary: this device is

hy-pass the detector ' nnd allows the bulk of the sample to

chelates formed^xith?'!,,'!?:-?”'^ suitable in the study of metal

involving chelates of noitvi . ^’cen used for quantitative work

the introduction of nu'orinent?*”"'^ ttfid fliiorinated derivatives. It is found that
flamcdetcctor which ic ' ■''^^'"*‘’*^P*t‘^^’'*t^^tIiminishcsthercsponseofthc
detector (Refs.' 2 and 1) cflcct observed xvith the electron capture

212

GAS CHROMATOGRAPHY IX, 2

_Jectr6n capture detector. Gases at near atmospheric pressure are
normally very good electrical insulators hut if they become ionised,_e^., by .
exposu re to (x~ or ^-radiation from a radioactive source, they will conduct an
'dectric^rrent. Most ionisation detectors are_based on measurement ot the'
increase in current (above that due to the background ionisation of the carrier
gas) which occuTs when a more readily ionised molecule appears in the gas
stream. The principles of operation of the various ionisation detectors have been
reviewed by J. E. Lovelock (Ref. 4). The electron capture detector differs from
other ionisation detectors in that it ‘exploits the recombination phenomenon, ~
being based on electron capture by compounds having an affinity for free
ele ctrons ; t he detector th us measjjres a decp^ e father than an jner-ease_m.,
^current. " ^ \

A j?-ray s ourceT commonly a foil containing) or ^ ^Ni) || s used to generate
‘slo^electrons by ionisation of the carrier gas (nitrogen preferred) flowing
Through the detector. These slow electrons migrate to the anode under a
pote ntial and give riseTo a steady base-line current. When an electron-capturing
gas (i.e., eluate molecules) emerges from the column and reacts with an electron,^
~ the net result is the replacement ot an electronJxv-a-negaUvejon ot much greater ’
mass w ith a corresno nding^d uction in current flow. O

The elec^n capture-detector isvery sensitive to certain molecules such as .
halogen-contaimn g compounds but insensitive to others such as hydrocarbons.
The response of the detector is clearly related to the electron affinity of the solute
molecules and, not surprisingly, it exhibits high sensitivity to fiuorinated fi-
diketonate complexes. The electron affinity of metal chelates appears to be a
function both of the nature of the metal ion and the extent of the halogenation in
the ligand. The detector is of great value in detecting ultra-t;ace amounts of
metals (Ref's)! ~

Element selective detectors. Many samples, e.g. tnoSe originating from
environmental studies, contain so many constituent compounds that the gas
chromatogram obtained is a complex array of peaks. For the analytical chemist,
who may be interested in only a few of the compounds present, the replacement
of the essentially non-selective type of detector (i.e. thermal conductivity, flame
ionisation, etc.) by a system which responds selectively to some property of
certain of the eluted species may overcome this problem.

The most common selective detectors in use at present respond to the presence
of a characteristic element or group in the eluted compound. The electron
capture detector, dealt with in detail above, comes in this category of selective
detectors. Similarly, the flame photometric detector, which is a modified form of
the flame ionisation detector (F.I.D.), has been specially developed for the
detection of phosphorus and sulphur compounds to the extent that its response is
10000 times greater for compounds containing these elements than for "
hydrocarbons. Another form of the F.I.D. is the thermionic detector which
employs a hydrogen flame burning at a jet with ah alkali metal salt tip and has a
selective sensitivity for compounds containing halogens, nitrogen, phosphorus
and sulphur (Ref 12). A particularly high degree of specific molecular
identification can also be achieved using on-line mass spectrometry or Fourier
transform infrared spectrometry, although these are normally employed for
orgamc compounds. The principles and applications of element selective
detectors have been reviewed (Ref 6).

The element specificity of atomic absorption spectrometry has also been used

213

IX. 3/4 QUANTITATIVE INORGANIC ANALYSIS

in conjunction with gas chromatography to separate anci determine organo-
mctallic compounds of similar chemical composition, c.g., lead alkyls in
petroleum : here lead is determined by A AS for each compound as it passes from
the gas chromatograph (Ref. 7).

IX, 3. PROGRAM MKD-TF.MPKRATURR G.AS CHROMATOGRAPHY.
Gas chromatognims arc usually obtained with the column kept at a constant
temperature. Two important disadvantages result from this isothermal mode
of operation:

1. Early peaks arc sharp and closely spaced (i.c. resolution is relatively poor in
this region of the chromatogram), whereas late peaks tend to be low, broad
and widely spaced (i.c., resolution is excessive).

2. Compounds of high boiling point arc often undetected, particularly in the
study of mixtures of unknown composition and uidc boiling point range; the
solubilities of the higher-boiling substances in the stationary phase arc so large
that they are almost completely immobilised at the inlet to the column
especially where the latter is operated at a relatively low temperature.

7Te above consequences of isothermal operation may be largely avoided by
using the technique of programmed temperature gas chromatography (PTGC) in
which the temperature of the whole column is raisctl during the sample analysis
(the \ arinlion of temperature with time may be linear or non-linear according to
the separation to be cflectcd). An alternative technique is chromathermography
in which a fixed temperature gradient is maimained down the column, the
column inlet being kept at the highest temperature and the outlet at the lowest.

Programmcd-tcmpcralurc gas chromatography pcrmit.s the separation of
compounds of a very wide boiling range moa' rapidly than by isothermal
operation of the column. The peaks on the chromatogram arc also sharper and
more uniform in shape so that, using PTGC. peak heights may be u.sed to obtain
.accurate quantitative analysis.

QUANTITATHT, ANALYSIS HV GLC. The quantitative dc-
fmination of a component in gas chromatography using dincrcntial-iypc
■'detectors of the type previously dc.scribed i s based upon mcasi i[CJUcnI-^f-dic
recorded p caiciirca or peak height ; the latter iMnorc suitable in the case of smal l
pea ks . or p eaksjyi 1 11" arrow b.and~\vadi]r In order that these quantities may be
“rcIatccTto ^he amount of s^dutirin the sample two conditions must prevail;
j («) the response of the detector-recorder system mii.sl be linear with respect to
,, the concentration of the solute;

Xl>) factors such as the rate of carrier gas flow, column temperature, etc., must be
kept constant or the clTcct of variation must be cliriiinated, c.g., by use of the
internal standard method.

Peak area is c ommonly used as a qimntitativ c-jueasu r c-oilai-parlJailar
component in the sample and can be measured by one of the following_
techniques : ~ —

v^^Hbnin ictry. The planiiiictcr is a mechanical d ejjicc-wliicIl-CnflhlCjJ liP
pe ^arca to be me asured by tracing the perimeter of t hejreakjhe method is
sld^V'bTrt'ca'nrglvc accurate rcsuli.s wTh experience in manipulation of the
planimeter.^ ^Accuracy and precision, however, ^lecrcasc as peak area diminishcs_ ^
2. Geometrical mctTiio3,s; imtie so-callcdtriangulation metfiods. tangents
are drawn to the inflection points of the elution peak and these two linc.s, together

214

Detector response

GAS CHROMATOGRAPHY IX, 4

Fig. IX, 3 Measurement of peak area by triangnlation

with the base line form a triangle (Fig. IX, 3) ; the area of the latter is calculated as
one-half the product of the base length times the peak height, the value obtained
being about 97 per cent of the actual area under the chromatographic peak when
this is Gaussian in shape (Ref. 8). . ■ .

The area may also be computed as the product of the peak height times the
width at half the peak height, i.e., by the height x width at half height method.
Since the exact location of the tangents (required for the triangulation method) to
the curve is not easily determined it is in general more accurate to use the method
based on width at half-height.

3. Integration by weighing. The chromatographic peak is carefully cut out
of the chart and the paper weighed on an analytical balance. The accuracy of the
method is clearly dependent upon the constancy of the thickness and moisture,
content of the chart paper, and it is usually preferable (unless an . automatic
integrator is available) to use geometrical methods.

4. Automatic integration. Integrators of this type may be divided into two
groups, viz., the mechanical type such as the ball and disc integrator, and the
more complex electromclype such as the digital int egrato r. These devices are
designed for attachment to the defector/recorder system so that integration of the
area may be carried out simultaneously with the recording of the chromatogram.
Electronic integrators give the best precision but are very expensive.

5. Data evaluation. It is, of course, necessary to correlate peak area with
the amount or concentration of a particular solute in the sample; this is usually
done by construction of a calibration graph of peak area versus amount of solute;
The calibration determinations must be carried out imder conditions which are
as similar as possible to those used in the chromatographic study of the sample.

Quantitative analysis using the internal standard method. The height and area of
chromatographic peaks’are'afreuled not only by the aiuount of sample but also
by fluctuations of the carrier gas flow rate, the column and detector temperatures
etc., i.e., by variations of those factors which influence the sensitivity and

215

IX, 5 QUANTITATIVE INORGANIC ANALYSIS

response ol'the detector. Ilic clTcct ofpeh variations can be eliminated by use of
the internal standard method in wliiclw known amount of a reference substance
is added to the sample to be analyst^ before injection into the column. The
rcquircmeni.s for an cflccti vc internal .standard (Section IV, 5) have been specified
as follows (Ref. 9):

(n) it sliouid give a completely resolved peak, but should be eluted close to the
components to be measured ;

(h) its peak height or peak area should be .similar in magnitude to those of the
components to be measured ;

(r) it .should be chemically similar to but not present in the original sample.

The. procedure comprises the addition of a constant amount of interna!
standard to ;i fi.xcd volume of several syntiieiic mi.vtums which contain vars'ing
known amounts of the component to be rietermined. The resulting mixtures are
chromatographed and a calibration curse is constructed of the per cent of
component in the mixtures against the ratio of component peak area/standard
peak area. The analysis of tlic unknown mixture is carried out by addition of the
same amount of internal standard to the specified volume of the mixture; from
the observed ratio of peak are.is the solute concentration is read off using the
calibration curve.

IX. 5. G.AS CHROM.ATOGRAPnV OF MKTAL CHI' L.ATES. Although
inorganic compounds arc generally not so volatile as arc organic compounds. ga.s
chromatography has been applied in the study of certain inorganic compounds
svhich possess tile requisite properties. If gas chromatography is to be used for
metal separation and quantitative analysis, liic types ofeompounds whichcan be
u-sed are limited to those that can be readily fonned in virtually quantitative and
easily reproducible yield. This feature, logcilicr with the requirements of
suflicicni volatility and thermal stability necessary for.succe.ssful gas chromato-
graphy, make neutral metal chelates the most favourable compounds for use in
metal analy.sis. //-Dikcione ligands, c.g.. acctylacctonc and the (luorinated
derivatives, trifluoroacetylacctonc (TEA) and hcxartuoroacelylacetonc (HFA)
form stable, vol.itilc chelates with aluminium, beryllium, chromiumflll) and a
number of other metal ions; it is thus possible tochromnlograpli a wide range of
metals as their //-diketonc chelates.

O O"

11 1

CFj-C-CI 1-C-CH, TFA anion

0 0 ~

II !

Cbj C — Cn — C — Cl '3 UFA anion

The number of reported applications to analytical dcicmiinalions at the trace

level appear to be few, probably the best known being the determination of

beiy'llium in various samples. The method generally involves the fonnation of

the volatile beryllium trifluoroacetylacctonatc chelate, its solvent extraction

into benzene xvith subsequent .separation and analysis by gas chromatography
(Ref. 10). 1 z j t

A number of important requirements must be met if gas chromatography is to
be successfully applied to metal analysis, and these will be briefly considered.

216

GAS CHROMATOGRAPHY IX, 5

1. Ease of formation of the metal compound. Chelating agents of the fi-
diketone type form complexes of high solvolytic stability by simple reactions
with metal ions. The reactions may take place in aqueous or non-aqueous media
and are pH dependent, a feature which provides a measure of selectivity when
mixtures of metal ions are being investigated. The solubility of these chelates in a
number of organic solvents permits their solvent extraction after preparation in
aqueous media, again providing enhanced selectivity, and yielding suitable
samples for injection into the chromatographic column.

2. Volatility. The most important requirement is thdt the metal com-
pound must be sufficiently volatile to be chromatographed in the gas phase. It is,
of course, not essential for the column to be operated at a temperature above the
boiling point of the compound and satisfactory elution will usually occur even if
the compound only possesses a vapour pressure of the order of a few mm at the
column temperature. The following types of metal compounds are volatile at
reasonably low temperatures: metal alkyls and alkoxides, metal carbonyls,
certain metal hydrides, metal cyclopentadienyls and related complexes with n-
acceptor ligands, and neutral metal chelates such as those formed with y?-
diketone ligands. When other factors, such as thermal stability of the compound,
are considered the choice of metal compound is limited to two major groups —
metal halides and metal y?-diketone chelates. An interesting feature of the latter
group is the markedly greater volatility of the fluorocarbon chelates (TFA and
HFA chelates) as compared with the corresponding acetylacetonates. Thus
chromium(III) hexafluoroacetylacetonate is eluted rapidly at quite low column
temperatures ranging down to 30 °C, whereas for a similar elution of
chromium(III) acetylacetonate the^column temperature required must be about
150 °C. The possibility of operating at lower column temperatures by using the
fluorocarbon chelates is clearly important in minimising any tendency to thermal
decomposition.

3. Thermal stability. An important criterion for quantitative work is that
the compounds should possess sufficient thermal stability to enable them to be
eluted without degradation ; it may be possible, however, to obtain quantitative
results even when thermal decomposition occurs provided that it is only slight
and that the extent of the decomposition is reproducible under the given
conditions.

Thermal degradation of the sample is indicated by the following observations :

(a) the presence of a residue in the injection port, although this may also be due

to incomplete vapourisation of the sample ;

(b) appearance of spurious chromatographic peaks ;

A more complete picture of the composition of the eluted material correspond-
ing to each gas chromatographic peak will, however, be obtained by collecting
and identifying the eluted material. A glass U-tube cooled in solid carbon dioxide
provides an adequate means of trapping metal chelates in the effluent ; the U-tube
is connected to the exit port of the column with a short length of Teflon tubing.
Identification of small effluent samples may be achieved using melting point
detennination, but physical techniques such as visible and ultraviolet spectros-
copy, infrared spectroscopy, etc., can often provide a more detailed analvsis of
the material. .

4. Solvolytic stability. In addition to being thermally stable, the metal
compounds must possess solvolytic stability particularly in relation to the liquid

217

JX, 6 QUANTITATIVE INORGANIC ANALYSIS

stationary pliasc in tlic column. If the molecules of the liquid phase function
effectively as ligands, solvolysis may occur through ligand substitution in the
metal complex. The compounds .should also not react chemically with the solid
stationary support or the materials of construction of the column. The reactivity
of the metal halides gives rise to a number of dilikuities :

(r/) the halidc.s arc easily hydrolysed and special precautions must be taken to
remove trace,s of moisture from the carrier gas;

{/)) at tiic elevated temperatures in the column the halides react with many of the
liquid stationary phases so that careful choice of the latter is required ;

Thus, despite the volatility of a number of metal halides their usefulness in gas
chromatography is limited by these considerations and the use of mcla) chelates
offers greater scope in metal analysis.

5. Health hanirds. Tlic effluent emerging from the gas chromatographic
instrument, if not trapped, will, of course, difi'ttsc into the laboraiors' atmosphere
and may constitute ;> heahli hazard. Tltis may happen if metals such as lead,
mercury ;md zinc, wliieh act as cumulative poisons, arc allowed to pass into the
laboratory atmosphere ;is \olatiic metal cornpleses. To prevent such atmo-
spheric contamination tiiecflliicnt stream should either be pa.ssed to a fume hood
or through a cold trap to remove the volatile metal compound. Appropriate
precautions must, of course, also be taken if the sample contains radioactive
material or when radioiscUopcs are employed in ionisation detectors.

IX. 6. Dirri'intiNATiON or ai.u.mimum rv g.as chromato-

CnAPHIC ANALYSIS OF ITS TRISfACLTVLACFTONATO)
COMPLEX. The purpose of this experiment is to illustrate the application of
gas chromatographic analysis to the quantitative detcnm'naiion of trace
amounts of metals as their chelate complc.xcs. The procedure described for the
determination of aluminium may ad.ipictl for the separation and de-
termination of aluminium and chromiumUH) as their acctylacctonate (Ref. 3).

S.amplc. Thesolvent e.xtraciion ofaluminium from aqueoussolution using
acetylacetonc (Ref. 1 1) can provide a suitable sample solution for ga.s chroma-
tographic analysis.

1 akc 5 cm’ of a solution containing about 1 5 mg of aluminium and adjust the
pil to between -1 and b. Equilibrate the solulion for 10 minutes with two
successive 5 cm’ portions of a solulion made up of equal volumes of ucetyiac-
ctonc (pure, redistilled) and A.R, chloroform. Combine the organic extracts.
Fluoride ion causes serious interference to the extraction and must be p-mvioii-sly
remcned.

Introduce a 0.30//I portion of the solvent extract into the gas chromatograph.
It is found that solutions of concentrations greater than 0.3X1 arc un.suitable as
they deposit solid and thus cause a blockage of the 1 pi microsyringc used for the
injection of the sample. The syring.c is flvishcd several times with the sample
solution, filled with the sample to the required volume, c.xcc.ss liquid wiped from
the tip of the needle and the sample injected into the chroniatogniph.

, Apparatus. A suitable instrument is the Pyc 104 Chromatograph equipped

with a flame ionisation detector, with an Autolab fi300 Digital Integrator linked
to a Westrex Teletype for printout. The use of a digital integrator is particularly
convenient for quantitative determinations, but other methods of measuring
peak area may be used (Section IX, 4).

218

GAS CHROMATOGRAPHY IX, 7/8

Pure nitrogen (white spot), at a flow rate of 40cm^min~‘, is used as carrier
gas. The dimensions of the glass column are 1.6 m length and 6 mm o.d., and is
packed with 5 per cent by weight SE-30 on Chromosorb W as the stationary
phase. The column is maintained at a temperature of 165 °C.

Procedure. Extract a series of aqueous aluminium solutions containing 5
to 25 mg aluminium in 5 cm^, using the procedure described above under Sample.
Calibrate the apparatus by injecting 0.30 pi of each extract into the column and
recording the peak area on the chromatogram. Plot a graph of peak area against
concentration.

Determine aluminium (present as its acetylacetonate) in the sample solution
by injecting 0.30 pi into the column. Record the peak area obtained and read off
the aluminium concentration from the calibration graph.

IX, 7. References

1. J. E. Schwarberg, R. W. Moshier and J. H. Walsh (1964). Talanta, 11 , 1213.

2. D. K. Albert (1964). ‘Comparison of Electron Capture and Hydrogen Flame
Detectors for Gas Chromatographic Determination of Trace Amounts of Metal
Chelates’, Anal. Chem., 36, 2034.

3. R. D. Hill and H. Gesser (1963). ‘An Investigation into the Quantitative Gas
Chromatographic Analysis of Metal Chelates using a Hydrogen Flame Ionisation
Detector’, J. Gas Chromatography, 1, 1 1.

4. J. E. Lovelock (1961). ‘Ionisation Methods for Analysis of Gases and Vapors’, Anal.
Chem.,33, 162.

5. W. D. Ross, R. E. Sievers and G. Wheeler Jr. (1965). ‘Quantitative Ultratrace
Analysis of Mixtures of Metal Chelates by Gas Chromatography’, Anal. Chem., 37,
598.

6. D. F. S. Natusch and T. M. Thorpe (1973). ‘Element Selective Detectors in Gas
Chromatography’, Anal. Chem., 45, 1 184A.

7. P. R. Ballinger and I. M. Whittemore (1968). Proceedings of the American Chemical
Society, Div. of Petroleum Chemistry, 13, [3], 1 33.

8. L. Condal-Bosch (1964). ‘Some Problems of Quantitative Analysis in Gas
Chromatography’, J. Chem. Educ., 41, A235.

9. D. Harvey and D. E. Chalkley (1955). Fuel, 34, 191 .

10. R. S. Barratt (1973). ‘Analytical Applications of Gas Chromatography of Metal
Chelates’, Proc. Soc. Analyt. Chem., 10 , 167.

11. I. M. Kolthoffand P. J. Elving(ed.) (1966). Treatise on Analytical Chemistry. Part II,
Vol. 4, p. 392. New York; Interscience.

12. R. C. Denney (1976). A Dictionary of Chromatography. London; Macmillan.

IX, 8. Selected bibliography

1. S. Dal Nogare and R. S. Juvet Jr. (1962). Gas-Liquid Chromatography. New York;
Interscience.

2. H. Purnell (1962). Gas Chromatography. New York; Wiley.

3. A. 1. M. Keulemans (1959). Gas Chromatography. 2nd edn. New York; Reinhold.

4. R. W. Moshier and R. E. Sievers (1965). Gas Chromatography of Metal Chelates.
Oxford ; Pergamon Press.

5. W. W. Brandt ( 1 963). ‘Gas Chromatography’, in Technique of Inorganic Chemistry, ed.
Jonassen and Weissberger, Vol. 3, p. 1 . New York; Interscience.

■ J. Trenchant (ed.) (1969). Practical Manual of Gas Chromatography. Amsterdam and
London; Elsevier.

219

PART D TITRIMETRY AND GRAVIMETRY

221

CHAPTER X , TITRIMETRICAMALYSIS

A. Theoretical Considerations

X, 1. TITRIMETRIC ANALYSIS. The term titrimetric analysis refers to
quantitative chemical analysis carried out by' determining the volume of a
solution of accurately known concentration which is required to react
quantitatively with the solution of the substance to be determined. The solution
of accurately known strength is called the standard solution; see Section X, 3. The
weight of the substance to be determined is calculated from the volume of the
standard solution used and the known laws of stoichiometry.

The standard solution is usually added from a burette. ITie process of adding
the standard solution until the reaction is just complete is termed a titration, and
the substance to be determined is titrated. The point at which this occurs is called
the equivalence point or the theoretical (or stoichiometric) end-point. The
completion of the titration should, as a rule, be detectable by some change,
unmistakable to the eye, produced by the standard solution itself (e.g., potassium
permanganate) or, more usually, by the addition of an^ auxiliary reagent, known
as an indicator. After the reaction between the substance and the standard
solution is practically complete, the indicator should give a clear visual change
(either a colour change or the formation of a turbidity) in the liquid being titrated.
The point at which this occurs is called the end-point of the titration. In the ideal
titration the visible end point will coincide with the stoichiometric or theoretical
end-point. In practice, however, a very small difference usually occurs; this
represents the titration error. The indicator and the experimental conditions
should be so selected that the difference between the visible end-point and the
equivalence point is as small as possible.

The term volumetric analysis was formerly used, but it has now been replaced
by titrimetric analysis, since it is considered that the latter expresses the process of
titration rather better, and the former may be confused with measurements of
volume, such as those involving gases. The reagent of known concentration is
called the titrant and the substance being titrated is termed the titrand. The
alternative name has not been extended to apparatus used in the various
operations: thus the terms volumetric glassware and volumetric flasks are still
retained, but it is probably. better to employ the expressions graduated glassware
and graduated flasks which are used throughout this book.

For use in titrimetric analysis a reaction must fulfil the following conditions:

1. There must be a simple reaction which can be expressed by a chemical
equation, the substance to be determined should react, completely with the

reagent in stoichiometric or equivalent proportions.

223

X, 2 QUANTITATIVE INORGANIC ANALYSIS

2. The reaction should be practically instantaneous or proceed with very great
speed. (Most ionic reaclions satisfy this condition.) In .some ca.scs the addition ofa
catalrst increa.scs the speed of a reaction.

3. There must be a marked change in free energy leading to alteration in some
physical or chemical property of the solution at the equivalence point.

4. An indicator should be available which, by a change in physical properties
(colour or formation of a precipitate), should sharply define the end point of the
reaction. (If no visible indicator is available for the detection of the equivalence
point, the latter can often be determined by following during the course of the
titration: (o) the potential between an indicator electrode and a reference
electrode (potcntiomefric titration, see Chapter XIV); (h) the change in electrical
conductivity of the solution (conductometric titration, see Chapter XV); (c) the
current which passes through the titration cell between an indicator electrode
(c-g.. the dropping mercury electrode) and a depolarised reference electrode (e.g.,
the saturated calomel electrode) at a suitable applied c.m.f. (ampcromelric
titration, see Chapter XVU); or (</) the change in absorbance of the solution
(spcctrophotomcfric titration, see Section X VUI, 39).)

Titrimctric methods are. as a rule, susceptible of higli precision (1 part in 1000)
and PO.SSCSS several advantagc.s. wiicrevcr applicable, over gravimetric mclhod.s.
They need simpler apparatus, and arc, generally, quickly performed; tcdiou.s and
difficult separations can often be avoided. Tiie following arc required for
titrimetne analysis: (i) calibrated mca.suring vessels, including burettes, pipettes,
and measuring flasks (see Chapter HI); (ii) substances of known purity for the
preparation of standard solutions; (iii) a visual indicator or an imstrumcntal
method for detecting tlic completion of the reaction.

X, 1 CLASSIFICATION OF REACTIONS IN TI TRIMETRIC ANALYSIS.
The reactions employed in titrimetricanalysi.s fall into two main chesses:

(fl) Tlioscin which no change in o.xidaiion stale occurs: ihc.se arc dependent upon
the combination of ions.

(h) Oxidation-reduction reactions; these involve a change of oxidation state or,
otherwise expressed, a transfer of electrons.

For purposes of convenience, however, these two types of reactions arc divided
into four main clas.ses;

1. Neutralisation reactions, or acidimciry and alkalimetry. Tlicsc include the
titration of free bases, or those formed from 5alt,s of weak acids by hydrolysis, with
a standard acid (acidimetry). and the titration offrcc acids, or those formed by the
hydrolysis of salts of weak bnsc.s. with a standard ba.se (alkalimcfrj). "nicse
reaclions involve the combination of hydrogen and hydroxidcion.s to form water.

2. Complex formation reactions. Thc,sc depend upon the combination of
ions, other than hydrogen or hydroxide ions, to form a soluble, slightly
dissociated ion or compound, as in the titration of a solution of a cyanide with
silver nitrate {2CN' +Ag * ^:i{Ag(CN),)') or of chloride ion with mercury(ll)
nitrate solution (2C1 “ + Hg^ ^ HgCIj).

Eihylcncdiaminctcira-aceticacid, largely as the disodium salt EDTA. is a very
important reagent for complex formation titrations and, indeed. EDTA has
beomc one of the most important reagents used in titrimctric analy.sis. Tlic use of
metal ion indicators has greatly enhanced its value in titrimetry. Tlic subject is
discu.s.scd fully later in this cliaptcr (Part A.2).

3. Precipitation reactions. These depend upon the combination of ions to

224

TITRIMETRIC ANALYSIS X, 3

form a simple precipitate as in the titration of silver ion v/ith a solution of a
chloride (Section X, 29). No change in oxidation state occurs.

4. Oxidation-reduction reactions. Under this heading are included all
reactions involving change in oxidation number or transfer of electrons (Section
X, 3) among the reacting substances. The standard solutions are either oxidising
or reducing agents. The principal oxidising agents are potassium permanganate,
potassium dichromate, cerium(IV) sulphate, iodine, potassium iodate, and
potassium bromate. Frequently used reducing agents are iron(II) and tin(II)
compounds, sodium thiosulphate, arsenic(III) oxide, mercury(I) nitrate,
vanadium(II) chloride or sulphate, chromium(n) chloride or sulphate, and
titanium(III) chloride or sulphate.

X,3. STANDARD SOLUTIONS. A standard solution is one which contains
a known weight of the reagent in a definite volume of solution, and for many years
concentrations were expressed in terms of molarity (i.e., number of moles per litre)
and normality (i.e., number of equivalents per litre). With the adoption by the
International Union of Pure and Applied Chemistry of the mole as a base unit of
quantity with the definition,

‘The mole is the amount of substance which contains as many elementary units
as there are atoms in 0.012 kilogram of carbon-12. The elementary unit must be
specified and may be an atom, a molecule, an ion, a radical, an electron or other
particle or a specified group of such particles’,

the mole is no longer a unit of mass, but is one of amount of substance, and terms
such as gram-molecule, gram-ion, etc., are obsolete.

With the introduction of this definition came proposals that the terms
‘molarity’ (the number of moles of solute per litre of solution), equivalent weight,
and normality (the number of equivalents of solute per litre of solution), should be
abandoned. However, experience has shown that there are certain practical
advantages in retaining the use of the terms equivalent and normal solution, and
the latest lUPAC recommendations (Ref. 1) suggest the following definitions:

‘The equivalent of a substance is that amount of it which, in a specified reaction,
combines with, releases or replaces that amount of hydrogen which is
combined with 3 grams of carbon-12 in methane ^ ^CH^.’

In this definition, the amount of hydrogen referred to may be replaced by the
equivalent amount of electricity or by one equivalent of any other substance,' but
the reaction to which the definition is applied must be clearly specified.

Although the terms mole and equivalent as now defined refer to an amount of
substance, each definition does in fact refer to a specified mass of carbon-12, and
hence we can say for example

1 mole of Hg2Cl2 has a mass of 0.47208 kg
1 mole of Na2CO3-10H2O has a mass of 0.286004 kg
1 mole of H2SO4 has a mass of 0.098078 kg
1 equivalent of Na2CO3-10H2O has a mass of 0.143002 kg

1 equivalent of H2SO4 has a mass of 0.049039 kg

and it is therefore quite permissible to refer to weighing out one mole of a certain
reagent, because this refers to a definite mass of the substance.

A normal solution is defined as a solution containing one equivalent of a defined

'75

X, 3 QUANTITATIVE INORGANIC ANALYSIS

species per dm-* according lo the specified reaction, and a molar solulion as one
containing one mole of a defined species per dm-', i.c„ a concentration of 1 mol
dm”-'.

As already explained (Section III, 12). the term litre is accepted as a special
name for tlic cubic decimetre, but with the suggestion that the litre should not be
iKscd to express results of liigh precision (Ref. 2) the recommendations of Ref, 1 can
be summarised as follows:

1. Wherever possible, concentrations should be expressed in terms of moles per
cubic decimetre (mo! dm ” -* or mol I ” ’).

2. The symbol Af to signify mol dm " ' should be retained, but the term molarity
should he discontinued.

3. The use of the term equivalent, defined as above, and given in the appropriate
S! unit should be retained, as should likewise the term normality based on the
redefined equivalent.

The above definition of normal solution utilises (he term ‘equivalent'. This
quantity varies with the type of reaction, and, since it is difiicull to give a clear
definition of ‘equivalent’ which will cover all reactions, it is proposed lo discuss
(his subject in .some detail below. It often happens that the same compound
possesses dilTercnt equivalents in dilTercnt chemical reactions, Tlic situation may
therefore arise in which a solution has norma! concentration when employed for
one purpo.se. and a different normality when used in another chemical reaction.

Neutralisation reactions. The cquiuiknt of an acid is tliai mass of it which
contains I.(X)fl (more accurately l.ixnfi) g of replaccabie hydrogen. The
equivalent of a nionoproticacid. such as hydrochloric, liydrobrnmic, hydriodic,
nitric, perchloric, or acetic acid, is kkntical with the mole. A normal solution of a
monoprolic acid will therefore contain 1 mole per dm' of solution. The
equivalent ofa diprotic acid (c.g., sulphuric or oxalic add), or of a triprotic acid
(c.g., phosphoric acid) is likesvise:! and J respectively of the mole,

T)ic cquiralcnt of a liasc is that mass of it which contains one replaixable
hydroxyl group, i.c., I7.(X).S g of ionisabic hydroxyl; I7.00S g of hydroxyl are
equivalent to 1.008 g of hydrogen. The equivalents of sodium hjtiroxidc and
potassium hydroxide are the mole, of calcium hydroxide, strontium hydroxide,
and barium hydroxide Haifa mole.

Salts of strong bases and weak acids possess alkaline reactions in aqueous
soUuion because of hydrolysis (Section I!, 17), A mole of sodium carbonate, with
mctliyl orange as indicator, reacts with 2 moles of hydrochloric acid to form 2
moles of sodium chloride ; hence its equivalent is half a mole. Sodium tetraborate,
under similar conditions, also reacts with 2 rnole.s of hydrochloric acid, and its
equivalent is. likewise, Haifa mole.

Complex formation and precipitation reactions. Here the equivalent is the
mass of the substance which contains or reacts with 1 mole ofa univalent cation
(which is equivalent to I.OOSgofhydrogcnj.i mole ofa bivalent cation M'*.
i mole of a tri valent cation M ' . etc. For the cation, the equivalent is the mole
divided by the valency. For a reagent which reacts with tliis cation, the equivalent
is the mass of it which reacts with one equivalent of the cation. The equivalent ofa
salt in a precipitation reaction is the mole divided by the total valency of the
mietiiig ion. Thus the equivalent of silver nitrate in the titration ofcliloridcionis
the mole.

In n complex formation reaction the equivalent is most simply deduced by
writing down the ionic equation of the reaction. For example, the equivalent of

226

TITRIMETRIC ANALYSIS X, 3

potassiuin cyanide in the titration with silver ions is 2 moles, since the reaction is .

2CN-+Ag+'^[Ag(CN)2]-- ■

In the titration of zinc ion with potassium hexacyanoferrate(II) solution:

3Zn"++2K4Fe(CN)6 = 6K++K2Zn3[Fe(CN)6]2

the equivalent of the hexacyanoferrate(II) is one-third of the mole. For other
examples of complex formation reactions, see Sections X, 19-27 ; it is apparent
that in many complexation reactions it is preferable to work in moles rather than
equivalents. . . . ,’

Oxidation-reduction reactions. The equivalent of an oxidising or reducing
agent is most simply defined as that mass of the reagent which reacts with or
contains 1.008 g of available hydrogen or 8.000 g of available oxygen. By
‘available’ is meant capable of being utilised in oxidation or reduction. The
amount of available oxygen may be indicated by writing the hypothetical
equation, e.g., . , .

2KMn04 = K20-F2Mn0 + 50 ,

i.e., in acid solution 2KMn04 gives up 5 atoms of available, oxygen, which is
taken up by the reducing agent, hence its equivalent is 2KMnO4/10.' For
potassium dichromate in acid solution, the hypothetical equation is: ■

“ K.20-i-Cr203 -b 30

The equivalent is K2Cr207/6. This elementary treatment is limited in
application, but is useful for beginners.

A more general and fundamental view is obtained by a consideration of:,(a) the
number of electrons involved in the partial ionic equation representing the
reaction, and {b) the change in the ‘oxidation number’ of a significant element in
the oxidant or reductant. Both methods will be considered in some detail. '

In quantitative analysis we are chiefly concerned with reactions which take
place in solution, i.e., ionic reactions. We shall therefore limit our discussion of
oxidation-reduction to such reactions. The oxidation of iron(II) chloride by
chlorine in aqueous solution may be written: ’

2FeCl2 + Cl2 = 2FeCl3

or may be expressed ionically: r

2Fe2+-bCl2 = 2Fe^ + -t-2Cr

The ion Fe^^ is converted into ion Fe^+ (oxidation), and the neutral chlorine
molecule into negatively charged chloride ions Cr (reduction); the conversion of
Fe into Fe^'*^ requires the loss of one electron, and the transformation of the
neutral chlorine molecule into chloride ions necessitates the gain of two electrons.
This leads to the view that, for reactions, in solution, oxidation is a process
involving a loss of electrons, as in

Fe^'*' —e = Fe^'*'

and reduction is the process resulting in a gain of electrons, as in
Cl,+2e = 2Cl- ,

227

X, 3 QUANTITATIVE INORGANIC ANALYSIS

In the actual oxidation-reduction process electrons are transferred from the
reducing agent to the oxidising agent. This leads to the following definitions.
Oxidation i.s the process which results in the loss of one or more electrons by
atoms or ions. Reduction is the proress which results in the gain of one or more
electrons by atoms or ion.s. An oxidising agent is one that gains electrons and is
reduced ; a reducing agent is one that loses electrons and is oxidised.

In all oxidation-reduction processc.s (or redox processes) there will be a
reactant undergoing oxidation and one undergoing reduction, since the two
reactions are complementary to one another and occur simultaneously — one
cannot take place without the other. The reagent suffering oxidation is termed the
reducing agent or rcductanf, and the reagent undergoing reduction is ctillcd the
oxidising agent or oxidant. The study of the electron chtmgcs in the oxidant and
rcductanl forms the basis of the ion-electron method for balancing ionic
equations. The equation is accordingly first divided into two balanced, partial
equations reprc.senting the oxidation and reduction respectively. It must be
rcmemlscrcd that the reactions take place in aqueous solution so that in addition
to the ions supplied by the oxidant and reduclant, the molecules of water HjO,
hydrogen ions H ’ , and hydroxide ions 01 1 ‘ are also present, and may be utilised
in balancing thcp.nrtial ionic equation. The unit change in oxidation or reduction
is a charge of one electron, which will be denoted by e. To appreciate the
principles involved. let us consider first the reaction between ironllll) chloride
and tinfll) chloride in aqueous solution. The p.artial ionic equation for the
reduction is:

Fc=‘ ( 1 )

and for the oxidation is;

Sn’^— . ( 2 )

The equations must be balanced not only with regard to the number and kind of
atoms, but also electrically, that is, the net electric charge on each side must be the
same. Equation ( 1 ) can be balanced by adding one electron to the left-hand side:

(F)

and equation (2) by adding two electrons to the right-hand side;

Sn"** :?iSn^M-2e (2')

These partial equations must then be multiplied by cocfiicicnts which result in the
number of electrons utilised in one rc.action being equal to those liberated in the
other. Tlius equation (F) must be multiplied by two, and wc have:

2Fc^^+2t>:ri2Fc=" (1")

Sn^'^ ::±Sn-*'’ -f2e (2")

Adding ( 1 ") and ( 2 "). we obtain:

2Fc3" -f2e^2Fc=" -fSn-*-* +2e

and by cancelling the electrons common to both sides, the simple ionic equation is
obtained:

2 Fc^++Sn*^ = 2 Fe*+ -1-811“

The following facts must be borne in mind. All strong elcctrolyte.s are

228

TITRIMETRIC ANALYSIS X, 3

completely dissociated; hence only the ions actually taking part or resulting from
the reaction need appear in the equation. Substances which are only slightly
ionised, such as water, or which are sparingly soluble and thus yield only a small
concentration of ions, e.g., silver chloride and barium sulphate, are, in general,
written as molecular formulae because they are present mainly in the
undissociated state.

The complete rules for the application of the ion-electron method may be
expressed as follows:

(a) ascertain the products of the reaction ;

(b) set up a partial equation for the oxidising agent ;

(c^ multiply each partial equation by a factor so that when the two are added the
electrons just compensate each other;

(e) add the partial equations and cancel out substances which appear on both
sides of the equation.

A few examples follow.

Reaction 1 : the reduction of potassium permanganate by iron(II) sulphate
in the presence of dilute sulphuric acid.

The first partial equation (reduction) is :

Mn04" — > Mn^^

To balance atomically, 8H is required : .

Mn04--f-8H+ -^Mn2++4H20
and to balance it electrically 5e is needed on the left-hand side:

Mn04 - + 8H + + 5c ^ Mn^ + -I- 4H2O
The second partial equation (oxidation) is :

Fe2+— »Fe^ +

To balance this electrically one electron must be added to the right-hand side or
subtracted from the left-hand side:

Fe^'*’ :^Fe^''‘ +e

Now the gain and loss of electrons must be equal. One permanganate ion utilises
5 electrons, and one iron(II) ion liberates 1 electron; hence the two partial
equations must apply in the ratio of 1:5.

Mn04 “ + 8H + -f 5e Mn^ + + 4H2 O

_ _ 5(Fe^+-Fe3++e)

Mn04“+8H+-(-5Fe^+ = Mn^ + 5Fe^ + + 4 H 2 O

Reaction II: the interaction of potassium dichromate and potassium
iodide in the presence of dilute sulphuric acid. .

Cr O > Cr^'*'

Cr207^'' + 14H+ — ^•2Cr^++7H20.

T 0 balance electrically, add 6e to the left-hand side :

Cr207^" + 14H++6e^2Cr^++7H20 ,

229

X, 3 QUANTITATIVE INORGANIC ANALYSIS

Tiic various stages in the dciiuclion of tlic second partial equation arc

r—h

2r -> I:

2r l2 + 2c

One dichroinatc ion uses 60 , and two iodide ions liberate 2{’; hence the two partial
equations apply in the ratio of 1 :3;

Cr, 0 ,‘* -fiqir +6ee:i2CVM-7H,0
3(21“ cil.-r2e)

or Cr^T= -’T 1 4H'*Tfd~~=^'2Cr^™

\Vc can now apply our knowledge of partial ionic equations to the subject of
equivalents. The standard o.xidation-rcduction process is H -fc. where e
represents an electron per atom, or the A vogadro's number of electrons per mole.
If we know the change in the numl>cr of electrons per ion in any oxidation-
reduction reaction, tlic equivalent may be calculated. Tlie equivalent of an oxid.ant
or a reductant is tlic mole divided by the number of electrons which I mole of the
substance gains or loses in the reaction, e.g.:

MnO^" f-Sir 4- SceetMrr* +4H.-0. Eq. = MnO^'/.^ = KMnO^/5.
Cr,0:- ' -t- 14tr + f)i’ ee:2Cr’ ' 4 711,0. Eq. r. Cr,0 ,- ' /6 <= K:Cri 07 / 6 .
Ec- ‘ Fc’ ‘ + c. Eq. Fe' \ 1 FcSO^ 'l ■

CjOr ■■ =e:2C0: 4-2c. hq. C,04'".^2 « H,C,0*/2.

SOj^' 4 -M 20 =:e:S 0 /- +213' 4- 2c. Eq. =- SO.,‘'/2 = Na^SOj/l

For convenience of reference the ptirtial ionic equations for a number of
oxidising and reducing agents arc collected in Table X. 1.

Tahte X, I. Ionic equations for use In the calculation of the cqiihalcnts of
oxidising and reducing agents

Subsianrc

ParJinI ionir equation

OXIDANTS

Potassium pcrmcmpaiiatc (add)
Potassium pcrmiinpanatc
(neutral)

Pomssium permanpanalc
(strongly .alkaline)

Ccrium(I V) f idpliate

Pot.assiuni didiromatc

Chlorine

Bromine

Iodine

Iron(I It) chloride
Potassium brom.ilc
Pola.ssium iodale
Sodium hypochlorite
llydtopen peroxide

MnOU +Sir -* SedtMn-* •) -Ul-O
MnO, ' 4 ;i 1 ,0 + .V ei .MaO. + -SO! I

-MnO,

Cc‘

* ■4’ t*

riCc”

, 0 -*'

a- Hit*

4 (s*

=i2Cr’*+7tLv

Cl.

+:•■

^2cr

Hr'

+ 2c

s+2Hr'

I.'

+ 2i-

=+2l‘

I-c'

‘ -l-r

c+Fc’ ■

BrOj'

■ +611*

4 6 i'

i^iBc ■ + 3 II 3 O

to,

■ +611*

46e

cif +3li;0

CIO

• 411,0

4 2t‘

ciCr +2011"

U.O

,+ 2 ir

+ 2c

;?S2U,0

230

TITRIMETRIC ANALYSIS X, 3

Substance

Partial ionic equation

Manganese dioxide
Sodium bismuthate
Nitric acid (cone.)
Nitric acid (dilute)

MnOj + 4 H-" + 2 e Mn^ + + 2H2O
BiOj - + 6 H+ + 2 e Bi^ + + 3H2O .
NO3" + 2 H+ +'e :^N02 +H2O
NO3-+4H+ +3e;^N0+2H20

REDUCTANTS

Hydrogen

Zinc

Hydrogen sulphide
Hydrogen iodide
Oxalic acid
Iron(n) sulphate
Sulphurous acid
Sodium thiosulphate
Titanium(III) sulphate
Tin(II) chloride
Tin(II) chloride (in presence of
hydrochloric acid)

Hydrogen peroxide

Zn:;iZn^++2e • ■
H2S:ri2H'" -ES+2e
2HI^l2+2H++2e
C 204 ^“:;± 2 C 02 + 2e
Fe^^ ^Fe^'*' +e

H2S03+H20^S04^''+4H++2e
2Sj03^':^S406^'-+2e
Ti3+ :;±Ti'‘'^+e
Sn^+;^Sn*^+2e

■Sn^-" +6Cr' ^SnCL^- +2e ,

- ■ H 2 O 2 ?i2H^ + 02 + 2e

The other procedure which is of value in the calculation of the equivalents of
substances is the ‘oxidation number’ method. This is a development of the view
that oxidation and reduction are attended by changes in valency and was
originally developed from an examination of the formulae of the initial and final
compounds in a reaction. The oxidation number (this will be abbreviated to O.N.)
of an element is a number which, applied to that element in a particular
compound, indicates the amount of oxidation or reduction which is required to
convert one atom of the element from the free state to, that in the compound. If
oxidation is necessary to effect the change, the oxidation number is positive, and if
reduction is necessary, the oxidation number is negative.

The following rules apply to the determination of oxidation numbers ;

1. The O.N. of the free or uncombined element is zero,

2. The O.N. of hydrogen (except in certain hydrides) has a value of+ 1,

3. The O.N. of oxygen (except in peroxides) is —2.

4. The O.N. of a metal in combination (except in hydrides) is usually positive.

5. The O.N. of a radical or ion is that of its electrovalency with the correct sigh
attached, i.e., is equal to its electrical charge.

6. The O.N. of a compound is always zero, and is determined by the sum of the
oxidation numbers of the individual atoms each multiplied by the number of
atoms of the element in the molecule.

The equivalent of an oxidising agent is determined by the change in oxidation
number which the reduced element experiences.' It is that quantity of oxidant
which involves a change of one unit in the oxidation number. Thus in, the normal
reduction of potassium permanganate in the presence of dilute sulphuric acid to a
Mn(II)salt:

+ 1 +T -8 +2+6-8 , .

KMn 04 — vMnS 04

the change in the oxidation number of the manganese is from + 7 to +2. The

231

X, 3 QUANTITATIVE INORGANIC ANALYSIS

equivalent of potassium permanganate is therefore I mole. Similarly for the
reduction of pota.ssium dichromatc in acid solution:

+ 2 ♦ j 2 ~ M + r. - (■

KXrjO, -^Cr,(SOJ.,

the change in oxidation number of two atoms of chromium is from + 12 to +6, or
by 6 unit.s of reduction. The equivalent of potassium dichromatc is accordingly ^
mole. In order to find the equivalent of an oxidising agent, we divide the mole by
the change in oxidation number per nwkaile which some key element in the
.substance undcrgoc.s.

The equivalent of a reducing agent is .similarly determined by the change in
oxidation number which the oxidised element suffers. Consider the conversion of
iron(ll)into iron(IH) sulphate*.

2(Fcs6j -* FcjfSOjj

Here the change in oxid;ition number per atom of iron is from + 2 to + 3, or by I
unit of oxidation, hence ilic equivalent of irontll) sulph.ite i.s 1 mole. Another
important reaction is the oxidation of oxalic acid to carbon dioxide and water:

^ 1 * 1 : a . .1 . 1

HjC,04 — 2CO,

The ciiungc in oxidation number of two atoms of carbon is from + 6 to -f 8, or by 2
units of oxidation. 'Hie equivalent of o.xalic jicid is therefore I mole.

In general, it may be stated:

(i) The equivalent of an element taking purl in an oxidation- retluction (redox)
reaction is the atomic mass divided by the change in oxidation number.

(ii) When an atom in any complc.x molecule suffers a cluinge in oxidation
number (o.xidation nr reduction), the equivalent of the substance is the mole
divided by the change in oxidation number of the oxidised or reduced element. If
more than one atom of the reactive element is present, the mole is divided by the
total change in oxidation number.

A u.seful summary of common oxidising and reducing agents, together with the
various transformations which they undergo is given in Table X. 2.

Table X, 2

S’lilwlancc Radical O.N.of

or 't.tfcc-

clcmcnl tiu''

itHobed element

CO.MMON OXIDI.SING AGHNT.S
KMnO.tacid) .MnOi' +7

KMnO^diculral) MnOj" +7

KMnOj (strongly

alkaline) MnO. ' -f7

KjCr.O, rr.O.= +6

HNOjfdil.) NOj •

HNOjIconc.) NO,- +S

Cl: Ci 0

Uc-dnetion

Nell

tle-

Gain

(irodticl

O.N.

crease

O.N.

in elec-
trons

.Mn=*

+ 2

MnO. fit
Mn* *■

-f4

MnO,= -

■(-6

Cr’*

-S ?

d-2

NO,

Cl-

-1

232

TITRIMETRIC ANALYSIS X, 3

Substance

Radical

element

involved

O.N. of

‘Effective’

element

COMMON OXIDISING AGENTS

Bfj

I 2

3HC1;1HN03

H 2 O 2

O 2

-1

Na 202

O 2

-1

KCIO 3

aoj-

+ 5

KBrOj

Br 03 -

+ 5

KIO 3

IO 3 -

+ 5

NaOCI

ocr

FeCh

Ce(SOj2

Ce"*-"

Reduction New Decrease Gain

product O.N. in in electrons

O.N.

Er-

-1'

l’

-1

ci-

-1

, 1

0 ^-

' -2

0 ^-

-2

ci-

-1

Br’

-1

ci-

-V

+ 2

Ce^ +

+ 3

COMMON REDUCING AGENTS

H 2 SO 3 or Na 2 S 03

SOj^*-

H 2 S

S^-

-2

-1

SnClj

Sn^+

+ 2

Metals, e.g., Zn

Hydrogen

FeS 04 ( or any

iron(II) salt)

Fe^*

+ 2

Na 3 As 03

AsOj^'

+ 3

H2C2O4

C204^-

+ 3

Ti2(S04)3

Ti3 +

S 04 ^-

+ 6

s°

1“

Sn''-^

■ +4

' 2

.Zn^+

+ 2

+ 1

Fe’+

+ 3

1 ■

AsOi®-

+ 5

CO 2

+ 4

Ti4 +

+ 4

We are now in a position to understand more clearly why the equivalents of
some substances vary with the reaction. We will consider two familiar examples
by way of illustration. A normal solution of ironfll) sulphate FeS04,7H20 will
have an equivalent of 1 mole when employed as a reductant, and j mole when
employed as a precipitant with aqueous ammonia. A solution of iron(II) sulphate
which is normal as a precipitant will be half normal as a reductant. Potassium
tetroxalate KHC204,H2C204,2H20 contains three replaceable hydrogen
atoms ; its equivalent in neutralisation reactions is therefore mole ;

KHC204,H2C204,2H20 + 3KOH = 2K2C2O4 + SHzO

As a reducing agent, a mole contains 2C204^ and the equivalent is accordingly
5 mole:

C2042--2e = 2C02

A solution of the salt which is 3N as an acid is 4 N as a reducing agent.

When a sequence of reactions is involved in a chemical process the reaction
which determines the equivalent is the one in which the standard solution is
actually used. Thus if sodium nitrate is reduced to ammonia with Devarda’s alloy
and the ammonia is titrated with standard acid, the equivalent of the sodium
nitrate is not determined by the reduction but by the reaction between ammonia
and the acid. Since the equivalent of ammonia NH3 is 1 mole, that of sodium

233

X. 4 QUANTITATIVE INORGANIC ANALYSIS

nitrate NaNO, is also I mole, bccti use 1 mole of NaNOj yields 1 moleofNHj,

X, 4. ADVANTAGlvS OF THE USE OF THE EQUIVALENT
SYSTEM. The most important advantage of tite equivalent system is that the
calculations of litrimetric analysis are rendered very simple, since at theend point
the number of equivalents of the substance titnited is equal to tlic number of
equivalents of the standard sohilion employed. We may write;

Number of cquivalcnt.s
tor''” '

Number of milli-cquivalcnts
Number of cm '

Hencc:numberofmilli-equivalcnts - number of cm' normality. Ifthcvolumcs
of solutions of two different substances A and B which e.vactly react with one
another arc I', cm' and I}, cm' respectively, then these volumes severally contain
the same number of equivalents or milli-cquivalcnts of A and B. Tlitis:

f', ttonnaluy^ = T^x normality, (I)

In practice l'^, F,. and normality, (the standard solution) arc known, hence
normality, (the unknown sohitionican be rcadil\’ calculated.

Extimplc t. How mans cm' of 0.2:V*hydroch!oric acid arc required to
neutralise 25.0 cm' of 0, 1 A’-sodium hydroxide'.'

Sub.slituting in equation ( I ). we obtain :

.X xO.2 = 2.5,0 X 0. 1, whence .v - 1 2.5 cm'

{■.sample 2- How many cm' of A'-hydrochlorie acid arc required to
precipitate completely I r, of .silver nitrate?

Tlic equivalent of ApNOji in a precipitation re.aclion is I mole or 169.89 g.
Hence I gof.AcNOj = I x l(X)b'l69.S9 ~ 5.886 milli-cquivalcnts.

Now number of milli-cquivalcnts of HCi - number of milli-cquivalcnts of
AgNOj;

.V X I ~ 5.8S6, whence .V — 5.90 cm'

Example .1. 25 cm' of an ironflll sulphate solution react completely with
.10.0 cm' of 0. 1 25iV-potassium permanganate. Calculate the strength of the iron
solution in grams of FCSO4 per dm'.

A normal solution of FeSO^ as a rcductani contains 1 mole per dm' or lSI.90g
per dm' (Table X, 2). Let the normality of the iron solution be Then:

25 X /I, = 30 X 0.125

or n, .10 X 0. 1 25/25 0. ! 50.V

Hence the solution wall contain 0. 1 50 x 1 5 1 .90 ~ 22,78 ,g EcSOj, per dm^.

£.\-ampli'4. What volume of0.127A' reaecnt is required for the preparation
of 1000 cm' of 0. 1 A' solution?

J/, X normality^ = I x normality,,

F, x0,127-.r, 1000 x 0.1

’i. = 1 000 X 0. J/0. 1 27 = 787.4 cm'

234

TITRIMETRIC ANALYSIS X, 5/6

Hence it is necessary to dilute 787.4 cm^ of Q.\21N solution to 1 dm^. Strictly
speaking it is not correct to, add 212.6 cm^ of water, because there is usually a
volume change on mixing. This change is so small, however, that diluted
solutions are often prepared by the addition of the calculated amount of water to
a measured volume of standard reagent.

X, 5. PREPARATION OF STANDARD SOLUTIONS. If a. reagent is
available in the pure state, a solution of definite normality is prepared simply by
weighing out an equivalent, or a definite fraction or multiple thereof, dissolving it
in the solvent, usually water, and making up the solution to a known volume. It is
not really essential to weigh out the equivalent (or a multiple or sub-multiple
thereof); in practice it is often more convenient to prepare the solution a little
more concentrated than is ultimately required, and then to dilute it with distilled
water until the desired normality is obtained. If Nj is the required normality,
the volume after dilution, Nj the normality originally obtained, and V 2 the
original volume taken, or — ^2 • The volume of water to

be added to the volume V 2 is {V^ — Fj) cm^ (compare Example 4 in Section X, 4).
The following is a list of some of the substances which can be obtained in a state of
high purity and are therefore suitable for the preparation of standard solutions;
sodium carbonate, potassium hydrogenphthalate, benzoic acid, sodium
tetraborate, sulphamic acid, potassium hydrogen iodate, sodium oxalate, silver,
silver nitrate, sodium chloride, potassium chloride, iodine, potassium bromate,
potassium iodate, potassium dichromate, and arsenic(III) oxide.

When the reagent is not available in the pure form as in the cases of most alkali
hydroxides, some inorganic acids and various deliquescent substances, solutions
of the approximate normality required are first prepared. These are then
standardised by titration against a solution of a pure substance of known
concentration. It is generally best to standardise a solution by a reaction of the
same type as that for which the solution is to be employed, and as nearly as
possible under identical experimental conditions. The titration error and other
errors are thus considerably reduced or are made to cancel out. This indirect
method is employed for the preparation, inter alia, of solutions of most acids (for
hydrochloric acid, the constant-boiling-point mixture of definite composition
can be weighed out directly, if desired), sodium, potassium and barium
hydroxides, potassium permanganate, ammonium and potassium thiocyanates,
and sodium thiosulphate. ’ .

X, 6. PRIMARY STANDARD SUBSTANCE& A primary standard
substance should satisfy the following requirements:

1. It must be easy to obtain, to purify, to dry (preferably at 1 10-120 °C), and to
preserve in a pure state. (This requirement is not usually met by hydrated
substances, since it is diflBcult to remove surface moisture completely without
effecting partial decomposition.)

2. The substance should be unaltered in air during weighing; this condition

implies that it should not be hygroscopic, nor oxidised by air, nor affected by
carbon dioxide. The standard should maintain its composition unchanged
during storage. , ®

235

X, 7 QUANTITATIVE INORGANIC ANALYSIS

3. The substance should be capable of being tested for impurities by
qualitative and other tests of known sensitivity. (The total amount of impurities
should not. in general, exceed 0.01-0.02 per cent.)

4. It should have a higli equivalent so that the weighing errors may be
negligible. (The precision in weighing is ordinarily 0. 1- 0,2 mg: for an accuracj' of
I part in 1000. it is necessary to employ samples weighing at least ca. 0.2 g.)

5. The substance should be readily soluble under the conditions in which it is
employed.

6. The reaction with the standard solution should be stoichiometric and
practically instantaneous. The titration error should be negligible, or ea.sy to
determine accurately by experiment.

In practice, an ideal primary standard is diflicull to obtain, and a compromise
between the above ideal requirements is usually necessary. The substances
commonly employed as primary standards arc: .Acid-base rcactions—sodium
carbonate Na,cbj. sodium tetraborate NajOiO^ potassium hydrogen-
phthalatc KlUCslljOj), constant-boiling-point hydrochloric acid, potassium
hydrogen iodate KH(!0,)-, bcnxoic acid H(C7H,Oj).

Complex formation reactions — silver, silver nitrate, sodium chloride, various
metals (c.g.. zinc, magnesium, copper, and speciro.vcopically pure manganese) and
salts, depending upon the reaction used.

Precipitation rcaclions—silvcr, silver nitrate, sodium chloride, potassium
chloride, and potassium bromide (prepared from poitissium bromatc).

Oxidation -reduction reactions — potassium dichromatc KjCr207, potassium
bromatc KBrOj, potassium iodate KlOj, potassium hydrogen iodate
KH(I03 )i, iodine K. sodium oxalate NajCjO^. arscnicflll) oxide AsjO^. and
pure iron.

Hydrated salts, as a rule, do not make good standards because of the difTiculiy
of efficient drying. However, tbose salts which do not clhorcsce, such as sodium
tetraborate NajD4O7.10H2O, and copper sulphate CuSOi.SH.O, are found by
experiment to be satisfactory secondary .standards. (See Ref. 1 1.)

A secondary standard is a substance which may be u.scd for standardisations,
and whose content of the active substance has been found by comparison against
a primary standard.

A. 1 THEORY OF ACID-BASE TITRATIONS

X, 7. NEUTRALISATION INDICATORS. The object of titrating, say. an
alkaline solution with a standard solution of an acid is the determination of the
amount of acid which is exactly equivalent chemically to the amount of base
present. The point at which this is reached is the equivalence point, stoichiometric
point, or thcorctic.al end-point: an aqueous solution of the corresponding salt
results. If both the acid and base arc strong electrolytes, the resultant solution will
be neutral and have a pH of 7 (Section U,l6): but ifeither the acid or the base is a
weak electrolyte, the salt will be hydrolysed to a certain degree, and the solution
at the equivalence point will be cither .slightly alkaline or slightly acid. Tlic exact
pH of the solution at the equivalence point can readily be calculated from the
ionisation constant of the weak acid or the weak base and the concentration of
the solution (see Section II, 18). Foranyactual titration the correct end point will

236

TITRIMETRIC ANALYSIS X, 7

be characterised by a definite value of the hydrogen-ion concentration, of the
solution, the value depending upon the nature of the acid and the base and the
concentration of the solution. . • .

A large number of substances are available, called neutralisation or acid-base
indicators, which possess different colours according to the hydrogen-ion
concentration of the solution. The chief characteristic of these indicators is that
the change from a predominantly ‘acid’ colour to a predominahtly ‘alkaline’
colour is not sudden and abrupt, but takes place within a small interval of pH
(usually about two pH units) termed the colour-change interval of the indicator.
The position of the colour-change interval in the pH scale varies widely with
different indicators. For most acid-base titrations we can therefore select an
indicator which exhibits a distinct colour change at a pH close to that obtaining
at the equivalence point.

The first useful theory of indicator action was suggested by W. Ostwald. All
indicators in general use are very weak organic acids or bases. Ostwald
considered that the undissociated indicator acid (HIn) or base (InOH) had a
different colour from that of its ion. The equilibria in aqueous solution may be
written;

HIn^H++In- (1)

and InOH^OH'+In+ (1')

unionised ionised

colour colour

If the indicator is a free amine 'or substituted amine, the equilibrium is : ' ‘

In + H20^0H"+HIn+ - (1")

Let us consider an indicator which is a weak acid. In acid solution, i.e., in the
presence of excess of H'*' ions, the ionisation will be depressed (coinrhon-ion
effect) and the concentration of In “ will be very small ; the colour will therefore be
that of the unionised form. If the medium is alkaline, the decrease of [H"*"] will
result in the further ionisation of the indicator; [In”] increases and the colour of
the ionised form becomes apparent. By applying the law of mass action, we '
obtain:

X flin- _ [H^] X [In-] Vh^ • yin- ..

%in [HIn] yHin ■" (2)

and

[H-^]

[HIn]

[In”]

xKi„x

yu^-yin-

[Un-ionised form] Vnin

Ft ^ TF T ^ ^

[Ionised form] Jh* . yj^-

(3)

where is the ionisation constant of the indicator. If the activity coefficients are
assumed to be unity— a not entirely justifiable assumption, as will be evident
Irom the ensuing discussion— equation (3) reduces to the simplified ‘con-
centration form’:

[H+] =

_ [HIn]

[In ]

[Un-ionised form]
[Ionised form]

(3')

237

X, 7 QUANTITATIVE INORGANIC ANALYSIS

The actual colour of the indicator, which depends upon the ratio of the
concentrations of the ionised and iin-ioniscd forms, is thus directly related to the
hydrogen-ion concentration. Equation (3') (the simplified or ’cla-ssicar form) may
be written logarithmically:

(4)

For an indicator which is a weak base an exactly analogous expression to (3')
may be deduced, wliich in its simplified form is:

(OH

UnOH]

(5)

where is now the corresponding base dissociation constant. This may be
written;

(in--

X'„.{In'OHJ

( 6 )

since f.'„, (H * ] x (OH ■“ ) {approximately)

Thesimple O.stwald theory of thecolourchangeofindicators requires revision,
but the modified \icws of indicator action lead to equations simihar to the above.
The colour changes arc believed to be due to structural changes, including the
production of quinonoid and resonance forms; these nitiy be illustrated by
reference to phcnolphthalcin, the changes of which arc characteristic of all
phtlmlcin indicators: sec the formulae (I-IV) given below. In flic presence of
dilute alkali the lactone ring in (1) opens to yield fll), and the triphcnylcarbinol
structure (11) undergoes loss of water to produce the resonating ion (III) which is
red. If phcnolphthalcin is treated with excess of concentrated alcoholic alkali the
red colour first produced disappears owing to the fonnation of (IV).

■•n

OH HO
OH

(

coo

(Ilf

OH *0
OH

- . . , - ..... .
r ^ i ] 11 L p

C(KJ

v-y

coo*

lUli

COO'

Tile Bronsted concept of acids and bases makes it unnecessary to distinguish
between acid and base indicators: emphasis is placed upon the charge types of the
acid and alkaline forms of the indicator. Tlie equilibrium between the acidic form
In^ and the basic form Inn ma}’ be expressed as:

238

. TITRIMETRIC ANALYSIS X, 7

InA^H++InB ; .

and the equilibrium constant as;

= , ( 8 )

^InA

The colour of an indicator, as perceived by the eye; is determined by the ratio of
the concentrations of the acidic and basic forms. This is given by ;

[Ipa] _ X Tine (9)

[Ins] ■ ,

where and yj,,^ are the activity coefficients of the acidic and basic forms of the
indicator. Equation (9) may be written in the logarithmic form :

pH = -logflH* = pf^in + iog|^+Iog^ (10)

UPa] Tiha

The pH will depend upon the ionic strength of the solution (which is, of course,
related to the activity coefficient— see Section 11, 3). Hence when making a colour
comparison for the determination of the pH of a solution not only must the
indicator concentration be the same in the two solutions but the ionic strength
must also be equal or approximately equal. The equation incidentally provides
an explanation of the so-called salt and solvent effects which are observed with
indicators. The colour-change equilibrium at any particular ionic strength
(constant activity-coefficient term) can be expressed by the modified equation ;

pH = pK',„ + log^ ’ , '

[Ipa] (11)

where pK'in is termed the apparent indicator constant.

The value of the ratio' [InB]/[InA] (i.e., [Basic form]/[Acidic form]) can be
determined by a visual colour comparison or, more accurately, by a
spectrophotometric method. Both forms of the indicator are present at any
hydrogen-ion concentration. It must be realised, however, that the human eye has
a limited ability to detect either of two colours when one of them predominates.
Experience shows that the solution will appear to have the 'acid’ colour, i.e., of
In^, when the ratio of [Iha] to [Ing] is above approximately 10, and the ‘alkaline’
colour, i.e., of Ing, when the ratio of [Ing] to [InJ is above approximately 10. Thus
only the ‘acid’ colour will be visible when [InAj/IIns] > 10, the corresponding l imi t
of pH given by equation (1 1) is ;

pH = pH:'.„-l

only the alkaline colour will be visible when [lnB]/[InA] > 10, and the

corresponding limit of pH is :
pH = pK',„ + l

The colour-change interval is accordingly pH = pii:'i„±l, i.e., over approxi-
mately two pH units. Within this range the indicator will appear to change from
one colour to the other. The change will be gradual, since it depends upon the
ratio of the concentrations of the two coloured forms (acidic form and basic form).

239

X, 7 QUANTITATIVE INORGANIC ANALYSIS

Und.sil m-><l Kx ^ !

Crrv'lfnl

QiiifliSJ -.f reJ

TltitT'*'* H-'C

kIsWMv

IJO

^frT4•C^’C'H.4

n= Blue

C'=^(’nhnirlrss
O-'Ofani'c
OB^Oraiire-haswn
P ••■ Purr’f

R==Kc4

RO^'Rcd-oranrf

Y'^YctUw

Mli

V i

“1w

- —

\yjMmiy

Sb.-

4-Nttru phrn.-.}

;

}l K

lVt'ra»'Cfi7s« I

Hof

...

Ntvtfi! ftU

lil

1“ j'Mf;x‘r‘'r

. !

. Vt^-v

r'.H'x'cr-t

Th)rt4 iV—

.. j ...

AK'j'n >i'r K

j 1 fSs\<Nv>NO!t 1

i. J

S' tn-’rr

- L .. ! .!

}

! L -1 ■ 11 ! ! ! i ■■ t I

n I ’ ? -i 5 6 7 y '» Ki !l 12 1.1

When the pM of the solution is equal to the apparent di.ssc>eialion constant of the
indicator pK'i„- tl'c ratio [In^| to [ln|,l becomes equal to 1, and the irulirator will
have a colour due to an equal mixture of the ‘acid’ and 'alkaline' forms. This is
sometimes known as the rnhldk' tint of the indicator. Thi.s applies .strictly only if
the two colours arc of equal intensity. If one form is more intensely coloured than
the other or if the eye is more .sensitive to one colour than the other, then the
middle tint will be slightly displaced along the pH range of the indicator.

Talilc X. 3. C.'olour changes anti pH rnnge of certain indicators

Indlealor

Oicmlcal name

pH range

Colour in

Arid

fvolntlon

Colour In
Alkaline
Solution

pX'in

Orilliant cresyl blue
(acid)

Amino-diclhylamino-

niclhy|.diphcna7onium

chloride

00-1.0

Ked-

oranse

Blue

—

Crcsol red (acid)

1 -Crcvotjulplionf-phlbalein

0 2-1.8

Red

Yeilovi

Quinaldinc red

t-(/i-Dimeih>l-amino-

phcn>l-clhylciK)-quinplinc

clhiodide

1.5--2.5

Colon rlcsc

Red

Thymol blue (acid)

Thym r>l -sill ph onc-phtlialei n

t.2-2.R

Red

Yellow

1.7

m-CrcsoI purple

ni-Crcsotsulptionc-phtli.atein

1.2 -2, 8

Red

Yellow

—

Pentamethoxy red

2,-t.2’.4'2'-!Vmanielhovy
tripheny) carbinol

1. 2-1.2

Red-

violet

Colourless

240

TITRIMETRIC ANALYSIS X, 7

Indicator.

Chemical name

pH range

Colour in

P^^'ln

Acid

Alkaline

Solution

Tropaeolin 00

Diphenylaminorp-
benzene-sodium sul-

1.3-3.0

Red

Yellow

—

phonate

Bromo-phenol blue

Tetrabromophenol-

sulphone-phthalein

3.0-4.6

Yellow

Blue

4.1

Methyl yellow

Dimethylamino-azo-

benzene

2.9-4.0

Red

Yellow

3.3

Ethyl orange

Dimethylamino-azo-

3.0-4.5

Red

Orange

—

Methyl orange

3. 1-4.4

Red

Orange

3.7

benzene sodium sul-
phonate

Congo red

Diphenyl-bis-azo- 1 -

naphthylaraine-4-sulphonic

acid

3.0-5.0

Blue

Red

Bromo-cresol green

Tetrabromo-m-cresol-

'3.8-5.4 '

Yellow

Blue

4.7

Methyl red ,

sulphone-phthalein
o-Carboxybenzene-azodi-
methyl-aniline ,

4.2-6.3

Red

Yellow

5.0

Ethyl red

4.5-6.5

Red

Orange

—

Propyl red

4.6-6.6

Red

Yellow

—

Chlorophenol red

4-Nitrophenol

Dichloro-phenol- ■ -

sulphone-phthalein

4.8-6.4

• Yellow

Red

6.1

4-Nitrophenol

5.6-7.6

Colourless

Yellow

7.1

Bromocresol purple

Bromophenol red

Dibromo-o-cresol-

sulphone-phthalein

5.2-6.8

Yellow

Purple

6.1

Dibromo-phenol-

sulphone-phthalein

5.2-6.8

Yellow

Red

Azolitmin (litmus)

—

5.0-8.0

Red

Blue .

—

Bromo-thymol blue

Neutral red

Dibromo-thymol-

sulphone-phthalein

6.0-7.6

Yellow

Blue

7.1

Amino-dimethyl-amino-
tolu-phenazonium chloride

6.8-8.0

Red

Orange -

Phenol red

Phenol-sulphone-

phihalein

6.8-8.4

Yellow

Red

7.8

Cresol red (base)

1-Cresolsulphone-

phthalein

7.2-8.8

Yellow

Red

8.2

1-Naphthol phthalein

1-Naphtholphthalein

7.3-8.7

Yellow

•Blue

8.4

m-Cresol purple

m-Cresolsulphone-

7.6— 9.2

Yellow

Purple

—

Thymol blue (base)

o-Cresol-phthalein

phthalein

Thyraol-sulphone-

phthalein

8.0-9.6

Yellow

Blue

8.9

Di-o-cresol-phthalide

8.2-9.8

Colourless

Red

—

Phenol-phthalein

Phenolphthalein

8.3-10.0

Colourless

Red

’ 9.6

Thymolphthalein

8.3-10.5

Colourless

Blue

9.3

Alizarin yellow R

p-Nitrobenzene-azo-

10.1-12.0

Yellow ,

Orange

—

Brilliant cresyl blue
(base)

salicylic acid

Amino-diethylamino-

methyl-diphenazonium

,10.8-12.0

Blue

red

Yellow

—

Tropaeolin 0

chloride

p-Sulphobenzene-azo-

11.1-12.7

Yellow

Orange

Nitramine

resorcinol

2,4,6-T rinitro-phenyl-

10.8-13.0

Colourless

Orange-

methyl-nitroamine

brown

241

X, 8 QUANTITATIVK INORGANIC ANALYSIS

Tabic X. 3 conlaiiis a selected list of indicators suitable for litrimclric analysis
and also for the colorimetric determination of pH. The colour-change intervals of
most of the various indicators listed in the table are represented graphically in
Fig. X, 1.

It is necessary to draw attention to the plf of various types of water which may
be encountered in quantitative analy.sis. Water in equilibrium with the normal
atmosplicre containing 0.03 per cent by volume of carbon dio,vidc has a pH of
about 5.7; very carefully prepared conductivity water has a pH close to 7; water
saturated with carbon dioxide under a pressure of one atmosphere has a pH of
about 3.7 at 25 'C. The analyst may therefore be dealing, according to the
conditions that prevail in the laboratory, with water having a pH between the two
c.xlremes pi 1 3.7 and pH 7. Hence for indicators which show their alkaline colours
at pH values above 4.5. the effect ofcarbon dioxide introduced during a titration,
either from the atmosphere or from the titrating solutions, must be seriously
considered. This subject is' discussed again later (Section X’, 12).

X. a PRKPARATIO.N OF INDICATOR SOLD HONS. As a rule the stock
solutions of the indicators contain 0.5- 1 g of indicator per litre of solvent. If the
substance is soluble in water, e.g., a .sodium salt, water is the solvent; in most
other cases 70-90 per cent ethanol is employed. It should now l>e .stated that the
synthetic indicators, particularly the sulplioncphthaleins and pliihaleins which
exhibit brilliant colour changes, may be u.scd with confidence in all those eases
where the older ones, largely natural products, were formerly employed.

Methyl orange. This indicator is encountered in commerce either as the
free acid or as the sodium salt.

Dissolve 0,5 g of the free acid in 1 litre of svater, and filler the cold solution if a
precipitate scparaic.s.

Dissolve 0.5 g of the sodium salt in I litre of water, add 15.2 env' of O.lAf-
hydrochloric acid, and filter, if necessary, when cold.

Methyl red. Dissolve 1 g of the free acid in 1 litre ofhol water, or dissolve in
600 cm’ of ethanol and dilute svith 400 cm’ of water.

Phenolphthnlcin, Dissolve 5 gofthctcngeni in 500 cm’ ofclhanolandadd
500 cm’ of water with constant stirring. Filter, if a precipitate forms.

Alternatively, dissolve 1 g of the dry indicator in 60 cm’ of 2-cthoxycth:mol
(ccllosolvc), b.p. 135 C, and dilute to 100 cm’ with distilled water: the loss by
evaporation is less with this preparation.

Thymolpbthalcin. Dissolve 0.4 g of the reagent in 600 cm’ of ethanol and
add 400 cm’ of water w ith stirring.

1-NaplilhoIphthaIcia Dissolve 1 g of the indicator in 500 cm’ of ethanol
and dilute with 500 cm’ of water.

Sulphoncphthalcias. These indicators are usually supplied in tlic acid form.
They arc rendered water-soluble by adding sudicient sodium hydroxide to
neutralise the potential sulphonic acid group. One gram of the indicator is
triturated in a clean glass mortar with the appropriate V>*intity ofO.lAf-sodium
hydroxide solution, and then diluted with water to 1 litre. The following volumes
of 0,1 A7 -sodium hydroxide arc required for I g of the indicators: bromo-phcnol
blue, 15.0 cm’; bromo-crcsol green, 14.4 cm’; bronio-crcsol purple. 18.6 cm’;
chloro-phcnol red, 23.6 cm’ ; bromo-thymol blue. 16.0 cm’ ; phenol red, 28.4cm’;
thymol blue, 21.5 cm’ icrcsol red, 26.2 cm’; meta-crcsol purple, 26.2 cm’,

Quinaldine Red. Dissolve 1 gin lOOcm’ of 80 percent ethanol.

242

TITRIMETRIC ANALYSIS X, 9

Methyl yellow, neutral red, and Congo red. Dissolve 1 g of the indicator in 1
litre of 80 per cent ethanol. Congo red may also be dissolved in water.

4-Nitrophenol. Dissolve 2 g of the solid in 1 litre of water.

Alizarin yellow R. Dissolve 0.5 g of the indicator in 1 litre of 80 per cent
ethanol.

Tropaeolin O and Tropaeolin OO. Dissolve 1 g of the sohd in 1 litre of
water. ■ '

Many of the indicator solutions are available commercially already prepared
for use. These should be bought from the actual chemical manufacturers, who will
usually supply full details as to the method of preparation, concentration of the
solution, etc.

X, 9. IVnXED INDICATORS. For some purposes it is desirable to have, a
sharp colour change over a narrow and selected range of pH; this is not easily
seen with an ordinary acid-base indicator, since the colour change extends over
two units of pH. The result may, however, be achieved by the use of a suitable
mixture of indicators; these are generally selected so that their pK'in values are
close together and the overlapping colours are complementary at an intermediate
pH value. A few examples will be given in some detail. .

(a) A mixture of equal parts of neutral red (0.1 per cent solution in ethanol) and
methylene blue (0. 1 per cent solution in fethanol) gives a sharp colour change from
violet-blue to green in passing from acid to alkaline solution at pH 7. This
indicator may be employed to titrate acetic acid with ammonia solution or vice
versa. Both acid and base are approximately of the same strength, hence the
equivalence point will be at a pH of ca. 7 (Section X, 15); owing to the extended
hydrolysis and the flat nature of the titration curve, the titration ' cannot be
performed except with an indicator of very narrow range.

(b) A mixture of phenolphthalein (3 parts of a 0.1 per cent solution in ethanol)

and 1-naphtholphthalein (1 part of a 0.1 per cent solution in ethanol) passes from
pale rose to violet at pH = 8.9. The mixed indicator is suitable for the titration of
phosphoric acid to the diprotic stage (K 2 = 6.3 x 10”® ; equivalence point at pH'
= ca.8.7). ■ .

(c) A mixture of thymol blue (3 parts of a 0.1 per cent aqueous solution of the
sodium salt) and cresol red (1 part of a 0.1 percent aqueous solution of the sodium
salt) changes from yellow to violet at pH = 8.3. It has been recommended for the
titration of carbonate to the hydrogen-carbonate stage.

Other examples are included in Table X, 4. The abbreviations p. = part. w.
= water, e = ethanol, Na = Na salt, are used. ■ ■ . - .

The colour change of a single indicator may also be improved by the addition
of a pH-sensitive dyestuif to produce the complement of one of the indicator
colours. A typical example is the addition of xylene cyanol FF to methyl orange
(1-0 g of methyl orange and 1.4 g of xylene cyanol FF in 500 cm^ of 50 per cent
ethanol): here the colour change from the alkaline to the acid side is greeii-*- grey
-+ magenta, the middle (grey) stage being at pH = 3.8.- The above is an example of
a screened indicator, and the mixed indicator solution is sometimes known as
screened’ methyl orange. Another example is the addition of methyl green (2
parts of a 0.1 per cent solution in ethanol) to phenolphthalein (1 part of a 0.1' per
cent solution in ethanol); the former complements the red-violet basic colour of
the latter, and at a pH of 8.4-8.8 the colour change is from grey to pale blue.- ■ ' >

243

X, IO/Jl/12 QUANTITATIVE INORGANIC ANALYSIS

TnWo X, 4. Some mixed Indicmors

IntSkatOT mi-vture

pit

Cotour chance

Composition

nromocreso! preen:

4,3

Orange

1 p.O.lfitNalin w.

methyl oriinpc

b'.ue-prcctr

1 p.O.’liinw.

lUomocrcsol green:

f..t

Pale preen — •

1 p, or,(N.iIinw,;

chloroplienol red

bine violet

I p.0I“..(,N’.i)inw.

Orotnolhvrnol blue:

7.2

Roec pint. —

I p. 0,1"; me,;

neutral red

preen

1 r,0.!%ine.

Bromothvmo! Wue:

7.3

bellow.' — •

1 p-O.UHNsii'"'.,

phenol red

violet

1 p. 0.!r.;iN’arin w.

Thymol blue;

Yellow — •

3 p 0 r;;(N.i)in w.

crcvol red

violcl

1 P LitNaJin w.

Thymol blue;

9.0

Vcllpa — •

! p 0 I”„ in jir,. c.;

phenolphlhalcin

violet

3p.Q,l’';.itt50'.,c.

'niymolphthalcin:

Colouticvv — •

1 p 0.1“^ me.;

phcnolphlludein

violet

1 p 0 17;, in «.

X, 10. universal OR MULTIPLE RANGE INDICATORS. By suitably
mixing curtain indicators changes in colour may occur over a considerable
portion of the pH range. Such mixtures arc usually called ‘universal indicators’.
They arc not suitable for qu.aniitativc titrations, but may be employed for the
determination of the appro.ximatc pH of a solution by the colorimetric method.
One such universal indicator is prepared thus; di.ssolvc 0,1 g of phcnolphthalein.
0.2 g of methyl red, 0..1g of methyl yellow. 0.4 gofbromothyrno! blue, and 0,5 g of
thymol blue in 500 cm^ of absolute ethanol, and add sodium hydroxide solution
until the colour is yellow. Tlie colour changes arc as follows: pH 2, red; pH 4,
orange; pH 6. yellow; pH 8, green; pH 10, blue.

Another recipe for the preparation of a universal indicator follows. Dissolve
0.05 g of methyl orange. 0.1 5 g of methyl red. 0.3 g of bromothymol blue, and 0.35
g of phcnolphthalein in 1 litre of 66 per cent ethanol. The colour changes arc: pH
up to 3, red; pH 4. orange-red; pH 5, orange; pH 6. yellow; pH 7, yellowish-green;
pH 8, grecni.sh-bluc: pH 9, blue; pH 10. violet; pH 11, reddi.sh-violet. Several
'universal indicators’ arc available commercially as solution.s and as test papers.

X, 11. NEUTRALIS.ATION CURVES. An insight into the mechanism of
neutralisation procc.sscs is obtained by studying the changes in the hydrogen-ion
concentration during the course of the appropriate titration, llie change in pH in
the neighbourhood of the equivalence point is of the greatest importance, as it
enables us to select an indicator which will give the smallest titration error. The
curve obtained by plotting pH as ordinates against the percentage of acid
neutralised (or tlic number of cm^ of alkali added) as abscissae is known as the
neutralisation (or, more generally, the titration) curve. This may be eraluatcd
experimentally by determination of the pH at various stages during the titration
by a potenliomctric method (Sections XIV, 16; 25). or it may be computed with
the aid of the theoretical principles that we have alrcadv studied. We shall, for the
present, adopt the latter method.

X, 12. NEUTRALISATION OF A STRONG ACID AND A STRONG
BASE. We shall assume that both the acid and the ba.se are completely
dissociated and that the activity cocllicicnts of tlic ions arc unity • order to

244

TITRIMETRIC ANALYSIS X, 12

Table X, 5. pH during titration of 100 cm^ of HCI wth NaOH of
equal concentration

Cm^ of NaOH
added

M solution
pH

O.IM solution
pH

O.OIM solution
pH

0.0

1.0

2.0

0.5

1.5

2.5

0.8

1.8

2.8

1.3

2.3

3.3

2.0

3.0

4.0

2.3

3.3

4.3

99.5

2.6

3.6

4.6

99.8

3.0

4.0

5.0

99.9

3.3

4.3

5.3

100.0

7.0

100.1

10.7

9.7

8.7

100.2

11.0

10.0

9.0

100.5

11.4

10.4

9.4

101

11.7

10.7

9.7

102

12.0

11.0

10.0

12.7

11.7

10.7

125

13.0

12.0

11.0

150

13.3

12.3

11.3

200

13.5

12.5

11.5

calculate the change of pH during the course of the neutralisation of the strong
acid and the strong base, or vice versa, at the laboratory temperature. For
simplicity of calculation we shall start with 100 cm^ of, say, M-hydrochloric acid
and add M-sodium hydroxide solution. The pH of M-hydrochloric acid is 0.
When 50 cm^ of the M base have been added, 50 cm^ of un-neutralised M acid
will be present in a total volume of 1 50 cm^.

[H+] will therefore be 50 x 1/150 = 3.33 x 10" \ or pH = 0.48
for75cm3 of base, [H+] = 25 x 1/175 = 1.43 x 10"\pH = 0.84
for90cm^ofbase,[H+] = 10x1/190 = 5.26 xl0-^ pH = 1.3
for 98 cm3 of base, [H+] = 2 X 1/198 = 1.01 X 10" ^ pH = 2.0
for 99 cm3 of base, [H+] = 1 X 1/199 = 5.03 X 10"3, pH = 2.3
for99.9cm3 ofbase, [H+] = 0.1 x 1/199.9 = 5.00x 10"^pH = 3.3

Upon the addition of 100 cm3 ofbase, the pH will change sharply to 7, i.e., the
theoretical equivalence point provided carbon dioxide is absent; the resulting
solution is simply one of sodium chloride.

With 100.1 cm3 ofbase, [OH"] = 0.1/200.1 = 5.00 x 10"^
pOH = 3.3 and pH = 10.7

With 101 cm3 ofbase, [OH"] = 1/201 = 5.00 x 10"3, pOH = 2.3
. and pH =11.7 ’ .

These results show that as the titration proceeds, initially the pH rises slowly but

addition of 99.9 and 100.1.cm3 of alkali, the pH of the solution rises

nuJlu ° Vicinity of the equivalence point the rate of change of

PWot the solution IS very rapid. , .

245

X, 12 QUANTITATIVE INORGANIC ANALYSIS

OT^-Tl'ilCTiclcm “■?■''' '“Si'*''’ ''"".‘'‘''I Ihrarasre in'™

reverse liiralfon of KB cm'> 'irfeK-',? ™ heyoiiil represems Ihe

hydrolysed solo,

praphically in Fin. X. 2. ^ ‘ presented

the cclliiva’lcnc%\,1n^^^^^^ iniercstcd in the chances of pH near

in Fie. X. 3, on which arL>-ilvntnrr-. on a larger scale

common indicators. * ' ‘^‘ohnir-changc intervals ofsorne of the

between pH 3 and loVni-ivV/'^*^''' /ai cfTcctive range

titration Sror ncphpbfc '

MethH olnnw \Wlf 'n^I'Oator i.s limited to 4.5-9.5.
been added, and tlic titration error 'n j n
most practical purposes; it is .i

solution until the indic-iinr ; ^'^'''^'**’1'^ to add .sodium hydroxide

titra.ionerro i a so ^ in the alkaline form. The

With nni W ? small with phenolphthalcin.

such indicatonsas^meSr^^^^ '’""P" ^^rtlicr limited to 5.5-8.5;

dioxide. In pmeaSt-uSd-^^ "‘h's'li do not contain carbon

and/or from Ihe atmosnlicre Tii darhonalc in Ihe sodium hydroxide

P ■ S'*-'* 's in equilibrium with carbonic acid, of

246

TITRIMETRIC ANALYSIS X, 13

Fig. X, 3 Neutralisation curves of 100 cm^ of HCI with NaOH of same concentration in
vicinity of equivalence point (calculated)

which both stages of ionisation are weak. This will introduce a small error when
indicators of high pH range (above pH 5) are used, e.g., phenolphthalein or
thymolphthalein. More acid indicators, such as methyl orange and methyl
yellow, are unaffected by carbonic acid. Kolthoff has calculated that the
difference in the amounts of sodium hydroxide solution used with methyl orange
and phenolphthalein is not greater than 0.15-0.2 cm^ of O.lM-sodium hydroxide
when 100 cm^ of O.lM-hydrochloric acid are titrated. A method of eliminating
this error, other than that of selecting an indicator with a pH range below pH 5, is
to boil the solution while still acid and to continue the titration with the cold
solution. Boiling the solution is particularly efficacious when titrating dilute (e.g.,

0.01 M) solutions.

X, 13. NEUTRALISATION OF A WEAK ACID WITH A STRONG
base We shall confine our attention to O.IM solutions; other concentrations
can be treated analogously. Let us study the neutralisation of 100 cm^ of O.IM-
acetic acid with O.lM-sodium hydroxide solution. The pH of the solution at the
equivalence point is given by (Section II, 18) :

pH = + ipK^ - ipc

= 7-M37-i(1.3) = 8.72

For other concentrations, we may employ the approximate mass action
expression:

[H+] X [CHaCOO-J/fCHaCOOH] = K, (1)

or [H+] = [CH3COOH] X KVfCHaCOO"]

or pH = log [Salt]/[Acid] -P pK„ (2)

247

X, 13 QUANTITATIVE INORGANIC ANALYSIS

The concentration of the salt (and of the acid) at any point is calculated from the
volume of alkali added, due allowance bcinit made for the total volume of the
solution.

The initial pM of O.lAf-acctic acid is computed from equation (1); the
dissociaiion of the acid is relatively so small that it may be neglected in expressing
the concentration of acetic acid. Hence from equation (1);

(ir jxlCHjCOO-MCH^COOlI] - 1.82x10-^

or [H ' 1 = V'T- W X 'l 0^' - ! .35 x 1 0 ' '
or pH =: 2.87

When 50 cm^ ofO.lAf-alkati have been added.

[Salt] - 50 .y 0.1/150 - .3.33x lO’’
and (Acid) - 50 x 0. 1 /1 50 3.33 x 10 " '

pH = log{3.33 X 10*T'3.33 x 10 "*)•)■ 4.74 -- 4.74

The pH values at other points on the titration curse are similarly calculated. .After
the equivalence point ha.s been passed, the solution contains e.xce.s.s of OH “ ions
whicli will repress the hydrolysis of tite salt; the pH may be assumed, with
sufficient accuracy for our purpose, to be that due to the excess of base present, so
that in this region the titration curve will almost coincide with that for O.lAf-
hydrochloric acid (Fig. X. 2 and Table X. 5). All the results arc collected in Table
X. 6. and arc depicted graphically in Fig. X. 4. Hie results for the titration of

Fig.X,4 Ncutr.aUsationcum>sofO.I.\/-accticacid.miIorO.Lt/-add(^. 1 x 10'’)
O.liV-sodiiini hydroxide (cnlculn(ed)

248

TITRIMETRIC ANALYSIS X, 14

Table X, 6. Neutralisation of 100 cm^ of O.lM-acetic acid (K„ = 1.82 x 10 *) and of
100 cm^ of0.1M-HA(X:^ = 1 x 10"’) with O.lAf-sodium hydroxide

Cm^ of 0.1M-NaOH
used

O.lM-acetic acid
pH

0.1M-HA(X„ = lxl0-’)
pH

2.9

4.0

3.8

6.0

4.3

6.5

4.7

7.0

5.7 .

8.0

99.0

6.7

9.0

99.5

7.0

9.3

99.8

7.4

9.7

99.9

7.7

9.8

100.0

8.7

9.9

100.2

10.0

100.5

10.4

101

10.7

110

11.7

125

12.0

150

12.3

200

12.5

lOOcm^ of O.IM solution of a weaker acid {Ka — 1 x 10~’) with O.lAf-sodium
hydroxide at the laboratory temperature are also included.

For O.lM-acetic acid and O.lM-sodium hydroxide, it is evident from the
titration curve that neither methyl orange nor methyl red can be used as
indicators. The equivalence point is at pH 8.7, and it is necessary to use an
indicator with a pH range on the slightly alkaline side, such as phenolphthalein,
thymolphthalein, or thymol blue (pH range, as base, 8.0-9.6). For the acid with
K^- 10 " the equivalence point is at pH = 10, but here the rate of change of pH
in the neighbourhood of the stoichiometric point is very much less pronounced,
owing to considerable hydrolysis. Phenolphthalein will commence to change
colour after about 92 cm^ of alkali have been added, and this change will occur to
the equivalence point; thus the end-point will not be sharp and the titration error
will be appreciable. With thymolphthalein, however, the colour change covers
the pH range 9,3-10.5; this indicator may be used, the end-point, will be more
sharp than for phenolphthalein, but nevertheless somewhat gradual, and the
titration error will be about 0.2 per cent. Acids that have dissociation constants
less than 10"’ cannot be satisfactorily titrated in O.IM solution with a simple
indicator.

In general, it may be stated that weak acids (K„ > 5x10"^) should be titrated
with phenolphthalein, thymolphthalein, or thymol blue as indicators.

X, 14. NEUTRALISATION OF A WEAK BASE ViTTH A STRONG
ACID. We may illustrate this case by the titration of 100 cm^ of O.lM-aqueous
ammonia = 1.8 x 10'^) with O.lM-hydrochloric acid at the ordinary
laboratory temperature. The pH of the solution at the equivalence point is given
by the equation (Section II, 18) :

pH = ipK^-ipRfc+ipc

= 7-2.37+1(1.3) = 5.28

249

X, 14 QUANTITATIVE INORGANIC ANALYSIS

For other conccnirations. the pFf may be calculated with suffident accuracy as
follows (compare previous section):

INH/JxfOirMNHji- A', (1)

or [Oir]-{Nn,l>cAVlNH/) (2)

or pOU - lop[Salt)/(B:isej + pA% (3)

or pH - pAl -pAV.~ log[SaItMIJasc) (4)

After the equivalence point has been readied, the .soimion contains excess of H *
ions, hydrolysis of the salt will be repressed, and the subsequent pH changes may
be assumed, with suHicienl accuracy for our purpose, to be those due to the excess
of acid present.

The results computed in the above manner are represented graphically in Fig.
X, .‘i; the results for the titration of 100 cm' of a 0.1 ,\f solution of a weaker base
(A,. - 1 X 10" ')nre also included.

Aoii .iUJcil. im'

Fig..\,5 NcutraILs.Kioncune«;oriOOcm’O.IAf-.aqiimisammonia(A'. = l.SxIO'")
andofO.I.tf-liaselAj == 1 x 10' ’)niili0.1.\f-lijtlrochloricadd.

It is clear iliat neither thymolphthaiein nor plienolphthalcin can be employed
in the titration of O.IA/-aqucou.s ammonia, llic equivalence point is at pH 5.3,
and it is necessary to use an indicator wiili a pH range on the slightly acid side (3-
6.5), such as methyl orange, methyl red. bromophcnol blue, or broniocrcsol green.
The last-named indicators may be utilised for the titration of all weak bases (A'^
> 5x lO”*) with strong acids.

For the weak base (A% = 1 x 10” '). bromo-pbcnol blue or methyl orange may
be used ; no sharp colour change will be obtained with bromo-crcsol green or with
methyl red, and the titration error will be considerable.

250

TITRIMETRIC ANALYSIS X, 15/16

X, 15. NEUTRALISATION OF A WEAK ACID WITH A WEAK
BASE . This case is exemplified by the titration of 100 cm^ of O.lM-acetic acid

= 1.8 X 10“®) with O.lM-aqueous ammonia (Kft = 1.8 x 10“®). The pH at the
equivalence point is given by (Section n, 18);

pH = ipK„ +ipK„ -ipKfc
= 7.0+2.37-2.37 = 7.0

The neutralisation curve up to the equivalence point is almost identical with that
using O.lM-sodium hydroxide as the base; beyond this point the titration is
virtually the addition of O.lAT-aqueous ammonia solution to O.lM-ammonium
acetate solution and equation (4) (Section X, 14) is applicable to the calculation of
the pH. The titration curve for the neutralisation of 100 cm® of O.lM-acetic acid
with O.lM-aqueous ammonia at the laboratory temperature is shown by the
dotted line in Fig. X, 4. The chief feature of the curve is that the change of pH near
the equivalence point and, indeed, during the whole of the neutralisation curve is
very gradual. There is no sudden change in pH, and hence no sharp end-point can
be found with any simple indicator. A mixed indicator, which exhibits a sharp
colour change over a very limited pH range, may sometimes be found which is
suitable. Thus for acetic acid-ammonia solution titrations, neutral red-
methylene blue indicator may be used (see Section X, 9), but on the whole, it is
best to avoid the use of indicators in titrations involving both a weak acid and a
weak base.

X, 16. NEUTRALISATION OF A POLYPROTIC ACID WITH A
STRONG BASE. The shape of the titration curve will depend upon the relative
magnitudes of the various dissociation constants. It is assumed that titrations
take place at the ordinary laboratory temperature in solutions of concentration
of O.IM or stronger. For a diprotic acid, if the difference between the primary and
secondary dissociation constants is very large (A'i/X 2 > 10000), the solution
behaves like a mixture of two acids with constants and K 2 respectively; the
considerations given previously may be applied. Thus for sulphurous acid,

= 1.7 X 10“ ® and K 2 = 1.0 x 10"^, it is evident that there will be a sharp change
of pH near the first equivalence point, but for the second stage the change will be
less pronounced, yet just sufficient for the use of, say, thymolphthalein as
indicator (see Fig. X, 4). For carbonic acid, however, for which Kj = 4.3 x 10“’
and K 2 = 5.6x 10“^\ only the first stage will be just discernible in the
neutralisation curve (see Fig. X, 4); the second stage is far too weak to exhibit any
point of inflexion and there is no suitable indicator available for direct titration.
As indicator for the primary stage, thymol blue may be used (see Section X, 13),
although a mixture of thymol blue (3 parts) and cresol red (1 part) (see Section X,
9) is more satisfactory; with phenolphthalein the colour change will be somewhat
gradual and the titration error may be several per cent.

It can be shown that the pH at the first equivalence point for a diprotic acid is
given by

[H+l= ‘

yj K, + c

Provided that the first stage of the acid is weak and that can be neglected by
comparison with c, the concentration of salt present, this expression reduces to

251

X, 16 QUANTITATIVE INORGANIC ANALYSIS

[iri- v/A',A'2,orpH-ip/^'i+:P^'2- .

With a knowledge of the pH at the stoichiometric point and also of the course
of the neutralisation curve, it should be an easy matter to select the appropriate
indicator for the titration ofany diprotic acid for which K./K. is at least If)^ For
many diprotic acids, however, the two dissociation constants arc loo close
tocethcr and it is not possible to dincrcnliiitc between the two stages. If is not
less than -about 10* % all the replaceable hydrogen may be titrated, c.g., sulphuric
acid (primary stace--a strong acid), oxalic acid, malonic. succinic, and tartaric

j^cicls.

.Similar remarks apply lo triprotic acids. Tlicsc may be illustrated bv reference
to ortliophosphoric acid, for winch K , " 7.5 x 10 ■ , K'j — 6.2 x 10 . and Kj
= 5 X 10"’l Here A, /A, 1.2 x lO' and A,/A, = 1.2 x lO', so that the acid

will behave as a mixture of three monoprolic acid.s with the dissociation
constants given above. Neutralisation proceeds almost completely to the end of
the primary stage before the .secondary .stage is appreciably afiected. and the
sccondarv stage proceeds almost to conipletion before the tcrliars stage is
apparent' The pH at the first cciiiivalcnce point is given approximately by (^pAj
+ IpA.) =■- 4.6, and at the second equivalence point by {ipAj-l- IpAji ^ 9.7: in
the very weak third stage, the curve is very flat and no indicator is available for
direct titration. The third equivalence point may be compvitcd approximately
from the equation (Section X, 13):
pH = -lpA«. + lpA„ - Ipr
== 7.0 + 6,15 -i (1.6)

= 12.-35forO.L\f HjP 04 .

252

TITRIMETRIC ANALYSIS X, 17

For the primary stage (phosphoric acid as a monoprotic acid), methyl orange,
bromo-cresol green, or Congo red may be used as indicators. The secondary stage
of phosphoric acid is very weak (see acid = 1 x 10“ in Fig. X, 4) and the only
suitable simple indicator is thymolphthalein (see Section X, 14); with
phenolphthalein the error may be several per cent. A mixed indicator composed
ofphenolphthalein (3 parts) and 1-naphtholphthalein (1 part) is very satisfactory
for the determination of the end-point of phosphoric acid as a diprotic acid (see
Section X, 9). The experimental neutralisation curve of 50 cm^ of O.IM-
orthophosphoric acid with O.lM-potassium hydroxide, determined by potentio-
metric titration, is shown in Fig. X, 6 .

There are a number of triprotic acids, e.g., citric acid with = 9.2 x 10“^^, Kj
= 2.7 X 10“^, X 3 = 1.3 X 10“®, the three dissociation constants of which are too
close together for the three stages to be differentiated easily. If K 3 > ca. 10" all
the replaceable hydrogen may be titrated; the indicator will be determined by the
value of K 3 .

X, 17. TITRATION OF ANIONS OF WEAK ACIDS (BR0NSTED BASES)
WITH STRONG ACIDS. ‘DISPLACEMENT TITRATIONS’. So far we
have dealt with titrations involving a strong base, the hydroxide ion, but
titrations are also possible with weaker bases (Bronsted bases), such as the
carbonate ion, the borate ion, the acetate ion, etc. Formerly titrations involving
these ions were regarded as titrations of solutions of hydrolysed salts, and the net
result was that the weak acid was displaced by the stronger acid. Thus in the
titration of sodium acetate solution with hydrochloric acid the following
equilibria were considered:

CH3 . COO" +H2O ^CHa . COOH + OH" (hydrolysis)

= H 2 O (strong acid reacts with OH" produced by

hydrolysis).

The net result thus appeared to be:

H^-t-CHaCOO" = CH3.COOH

or CHa.COONa+HCl = CHa.COOH-bNaCl

i.e., the weak acetic acid was apparently displaced by the strong hydrochloric
acid, and the process was referred to as a displacement titration. On the Bronsted
theory the so-called titration of solutions of hydrolysed salts is merely the
titration of a weak Bronsted base with a strong (highly ionised) acid. When the
anion of a weak acid is titrated with a strong acid the titration curve is identical
with that observed in the reverse titration of a weak acid itself with a strong base
(compare Section X, 13).

A few examples encountered in practice will now be considered.

Titration of borate ion with a strong acid. The titration of the tetraborate
ion with hydrochloric acid is similar. The net result of the displacement titration
IS given by:

B4O72- + 2 H+ -b5H20 = 4 H 3 BO 3

Boric acid behaves as a weak monoprotic acid with a dissociation constant of

253

X, J6 QIJANTITATIVE INORGANIC ANALYSIS

{ir j = '’<■ pJ’ ■■■ , . . .

Wilh a knowIcdi:c of flic pi I .’tl ilic stoichiomcinc poini and also of thccoura
of the neuUalisatiori curve, if should he ao easy muder to select the approprbit
inclic.itor for the lilraiion of any diprotic acid for which K,/A'; is at least KP.Fot
many diprotic aciils, liowcvcf. the two rlisvocialion constants arc too efe;
together and it is not possible ts» diirerctniaic hctsscen the two stapes. IfK. isaoi
less than about 1 H ’ \ all the rcplace.ahlc hydropen may be tit rated, c.e., sulphuric
acid (primary staec - a .stronsr acid), oxalic acid, malonic, snccinic. and tartatic
acids.

Simitar remarks .apply to triprotic acids. These may he illustrated by reference
to orthophfsphoric acid, for which h , ■' 7.5 It) 6.2 x 10*^. and kh

Ts 5 X 10"*^, Here K,/A'; 1.2 >• U)' and K; K > ~ 1,2 I0\ so that the add

will bchasc as a mixture of three morioproiic acids wiili the dis>od.rtic<n
constants given above. Nciittalijatioti proceeds almost compleidy to the end of
the primary stage before the secondary stage is appreciably affected, and It;
secondary' stare proceeds almost to ci>mp!etion before the tertiary slacj u
appatent. The pH at t!\e first etiuhalcnce point is given approxim.atcly byiipk',

■ jpTse) ‘hb. and at the second cquiv.'.icncc point by l*pfC > jpKj) ?■ 9.7; ia
the very weak third sl.ii’e. the cone is very (I.n and no indicator is as.iibHcfcr
direct titration. Tiic dnrd cqtiisaicnce fwiint tnay Ive computed approximaidy
from the equation (Section X, 13):

pH ;PA’. -t ;Pfs j “ipt
-- 7.0 + 6 l.f- 1(1,61
-- 12,35 forO.l.Vf HT’O.,.

Fir. X. 6 Titration ofSOaiH ofO.I.l/.HjPO, with fl.l.tf-KOH

252

TITRIMETRIC ANALYSIS X, 18

X, 18. CHOICE OF INDICATORS IN NEUTRALISATION RE-
ACTIONS. As a general rule it may be stated that for a titration to be feasible,
there should be a change of approximately two units of pH at or near the
stoichiometric point produced by the addition of a small volume of the reagent.
The pH at the equivalence point may be computed by means of the equations
given in Section n, 18 (see also below), the pH at either side of the equivalence
point (0.1-1 cm^) may be calculated as described in the preceding sections, and
the difference will indicate whether the change is large enough to permit a sharp
end point to be observed. Alternatively, the pH change on both sides of the
equivalence point may be noted from the neutralisation curve determined by
potentiometric titration (Sections XIX, 16; 25). If the pH change is satisfactory,
an indicator should be selected that changes at or near the equivalence point.

For convenience of reference, we shall summarise the conclusions already
deduced from theoretical principles.

Strong acid and strong base. For O.IM solutions or stronger, any indicator
may be used which has a range between the limits pH 4.5 and pH 9.5. With 0.0 IM
solutions, the pH range is somewhat smaller (5.5-8.5). If carbon dioxide is
present, the solution should either be boiled whilst still acid and the solution
titrated when cold, or an indicator with a range below pH 5 be employed;

pH 14
13
12
11
10
9
8
7
6
5
4
3
2
1
0

Fig.X,7 Titration of lOOcm® of O.lM-NazCOjTrithO.lM-Ha

Weak acid and a strong base. The pH at the equivalence point is calculated
irom the equation:

pH = 2pX;„ + ipX„ - |pc

pH range for acids with > 10“^ is 7-10.5; for weaker acids (K > 10“®)
the range is reduced (8-10). The pH range 8-10.5 will cover most of the examples

Indicator

ranges

^ Phenol -
$ phlhalein

Thymol blue

Bromo-'

Dhenol

jiue

3 Methyl

S orange

0 10 20 30 40 50 60 70 80 90 100
Add added, cm^

255

TITRIMETRIC ANALYSIS X, 19/20

A. 2 THEORY OF COMPLEXATION TITRATIONS

X, 19. INTRODUCTION. A; complexation reaction with a metal ion
involves the replacement of one or more of the co-ordinated solvent molecules by
other nucleophilic groups. The groups bound to the central ion are called ligands
and in aqueous solution the reaction can be represented by the equation:

M(H20)„-^L = M(H20)(„_i)L + H20.

Here the ligand (L) can be either a neutral molecule or a charged ion, and
successive replacement of water molecules by other ligand groups can occur until
the complex ML„ is formed; n is the coordination number of the metal ion and
represents the maximum number of monodentate ligands that can be bound to it.

Ligands may be conveniently classified on the basis of the number of points of
attachment to the metal ion. Thus simple ligands, such, as halide ions or the
molecules HjO or NH3, are monodentate, i.e., the ligand is bound to the metal ion
at only one point by the donation of a lone pair of electrons to the metal. When,
however, the ligand molecule or ion has two atoms, each of which has a lone pair
of electrons, then the molecule has two donor atoms and it may be possible to
form two coordinate bonds with the same metal ion; such a ligand. is said to be
bidentate and may be exemplified by consideration of the tris(ethylenedi-
amine)cobalt(III) complex, [Co(en)3]^'*'. In this 6-coordinate octahedral complex
of cobalt(III), each of the bidentate ethylenediamine molecules is bound to the
metal ion through the lone pair electrons of the two nitrogen atoms. This results
in the formation of three 5-membered rings, each including the metal ion; the
process of ring formation is called chelation. .

Multidentate ligands contain more than two coordinating atoms per molecule,
e.g., 1,2-diaminoethanetetra-acetic acid (ethylenediaminetetra-acetic acid,
EDTA) which has two donor nitrogen atoms and four donor oxygen atoms in the
molecule can be hexadentate.

In the foregoing it has been assumed that the complex species does not contain
more than one metal ion, but under appropriate conditions, a binuclear complex,
i.e., one containing two metal ions, or even a polynuclear complex containing
more than two metal ions may be formed. Thus interaction between Zn^ and
Cl~ ions may result in the formation of binuclear complexes, e.g., [Zn2Cl6]^'' in
addition to simple species such as ZnClj " and ZnCl4^ “ . The formation of bi- and
poly-nuclear complexes will clearly be favoured by a high concentration of the
metal ion; if the latter is present as a trace constituent of a solution, poly-nuclear
complexes are unlikely to be formed.

X, 20. STABILITY OF COMPLEXES. The thermodynamic stability of a
species is a measure of the extent to which this species will be formed from other
species under certain conditions, provided that the system is allowed to reach
equilibrium. Consider a metal ion M in solution together with a monodentate
ligand L, then the system may be described by the following step-wise equilibria,
in which, for convenience, coordinated water molecules are not shown :

M-bL^ML; if 1 = [ML]/[M1[L]

ML-bL^ML^; if^ = [ML2]/[ML][L]

ML(„_,,-bL^ML„; = [MLJ/[ML(„_3)][L]

257

X, }8 QUANTITATIVE INORGANIC ANALYSIS

likely to he cncoimtcrcci; this permits of the use of thymol blue, thymolplnhalcia,
or phcnolphtlialcin.

Weak base and stronR acid. Tire pH at the equivalence point is computed
from the equation:

pH ipArH.--lpK^-l-ipc

The pH range for bases with A.',, > 10" ' is 3--7, and for weaker bases (K^ > 10"*)
3-5. Suitable indicators will be methyl red, methyl orange, methyl yellow, bromo-
crcsol green, and bromo-phcnol blue.

Weak acid and sveak base. There is no sharp rise in the neutralisation curve
and, generally, no simple indicator can be used. The titration should therefore be
avoided, if po.ssiblc. ITe approxim.atc pH at the equivalence point can be
computed from the equation:

pH =- lpA'„ 4-lpA'«~lpAT

It is sometime.s possible to employ a mixed indicator (see Section X, 9) which
cxhibit.s a colour change over a very limited pH range, for example, neutral red-
methylene blue for ammonia .solution and acetic acid.

Folyprotic acids (or mixture.s of acids, with dissociation constants A',. A',,
and A'd and strong bast's. The first stoichiometric end point is given
approximately by;

pH = J(pA', + pA',)

The second stoichiometric ctid point is given approximately by;

pH -- J(pA', + pA',)

Anion of a ncak acid titrated nith a strong acid. The pH at the equivalence
point is given by:

pH “ ipA',,-ipA„-lpc

Cation of a ncak base titrated with a strong base. The pH at the
stoichiometric end-point is given by;

pH == ipA'., -ipA^-ipr

As a general rule, wherever an indicator docs not give a sharp end point, it is
advisable to prepare an equal volume of a comparison solution containing the
same quantity of indicator and of the final products and other components of the
titration as in the solution under te.st. and to titrate to the colour shade thus
obtained.

TITRIMETRIC ANALYSIS X, 21

and mercury(II) form strong complexes with I" and CN“ ions, but weak

complexes with F“. ■ . i.

3. Transition metal ions with incomplete d sub-shells. In this group both Class
A and Class B. tendencies can be distinguished. The elements with Class B
characteristics form a roughly triangular group within the Periodic Table, with
the apex at copper and the base extending from rhenium to bismuth. To the left of
this group, elements in their higher oxidation states tend to exhibit Class A
properties, while to the right of the group, the higher oxidation states of a given
element have a greater Class B character.

The concept of Hard and Soft acids and bases is useful in characterising the
behaviour of Class A and Class B acceptors. A soft base may be defined as one in
which the donor atom is of high polarisability and of low electronegativity, is
easily oxidised, or is associated with vacant, low-lying orbitals. These terms
describe, in different ways, a base in which the donor atom electrons are not
tightly held, but are easily distorted or removed. Hard bases have the opposite
properties, i.e., the donor atom is of low polarisability and high electronegativity,
is difficult to reduce, and is associated with vacant orbitals of high energy which
are inaccessible.

On this basis, it is seen that Class A acceptors prefer to bind hard bases, e.g.,
with nitrogen, oxygen and fluorine donor atons, whilst Class B acceptors prefer to
bind to the softer bases, e.g., P, As, S, Se, Cl, Br, I donor atoms. Examination of the
Class A acceptors shows them to have the following distinguishing features; small
size, high positive oxidation state, and the absence of outer electrons which are
easily excited to higher states. These are all factors which lead to low
polarisability, and such acceptors are called hard acids. Class B acceptors;
however, have one or more of the following properties: low positive or zero
oxidation state, large size, and several easily excited outer electrons (for metals
these are the d electrons). These are all factors which lead to high polarisability,
and Class B acids may be called soft acids.

A general principle may now be stated which permits correlation of the
complexing ability of metals: ‘Hard acids prefer to associate with hard bases and
soft acids with soft bases’. This statement must not, however, be regarded as
exclusive, i.e., under appropriate conditions soft acids may complex with hard
bases or hard acids with soft bases. ' . .

(b) Characteristics of the ligand. Among the characteristics of the ligand
which are generally recognised as influencing the stability of complexes in which
it is involved are (i) the basic strength of the ligand, (ii) its chelating properties (if
any), and (iii) steric effects. From the point of view of the analytical applications of
complexes, the chelating effect is of paramount importance and therefore merits
particular attention. ,

The term ‘chelate effect’ refers to the fact that a chelated complex, i.e., one
formed by a bidentate or a multidentate ligand, is more stable than the
corresponding complex with monodentate ligands: the greater the number of
points of attachment of ligand to the metal ion, the greater the stability of the
complex. Thus the complexes formed by the nickel(II) ion with (a) the
monodentate NH 3 molecule, (h) the bidentate ethylenediamine ( 1 , 2 -diaminoe-
thane), and (c) the hexadentate ligand ‘penten’ {(H 2 N-CH 2 -CH 2 ) 2 N-CH 2 -CH 2 -N
(CH 2 -CH 2 -NH 2 ) 2 } show an overall stability constant value for the ammonia
complex of 3. 1 X 10®, which is increased by a factor of about 10* ° for the complex
of hgand (h), and is approximately ten times greater still for the third complex

259

X, 21 QUANTlTATIVf- INORGANIC ANALYSIS

TIic equilibrium consuints K\. , . K arc rcfcrrci »

conslants, ' ‘'‘S' stcp-nisc stability

An nlfi»riinf it »»

iis follows;

constants.

An alternative way of expressing the equilibria is

;ML: //, =r [MLj/jMjlLj

^ML,: p, ... (ML,J/[MJ(L]=

= ML.; A* -

M + L:
M 4" 2L i
M -f nL z

1 he equilibrium constants ' n n ,woi ,

/i„ - A', K A% X . . . A‘„.

analytiwl cheim\-tr^\^ sS wnsicJerable importance in

the various complc.vcs formed bf a mefil in so - r of

invaluable in the study of

procedures such as solVcnl e\traciion ion analytical separation

20and 21). ^^traci.on, ,on exchange, and chromatography (Refs.

,brSy'siS;i“

I.C., electron acceptors. Class A rnenkV* ^

nqueous solution) towards the nloecnTr "" «rdcr of affinity (in

{no.ststablecomple,xcswi,h t > I -- and form their

Penodic Table (i.e., nitrogen oxvcen . h n atoms in the

much more readily with r th-m S r ' ^ia.ss B metals coordinate

most stable compIe.\cs with thc'sernn/i / solmion. and fonn their

(i.e., P. s. Cl). The Schwar7cnlnch ch<.<T donor atom from each group

ion acceptors : ^'‘t^'tl'cat.on dcllnes tha-e categories of metal

I. Cations with noble ms m r

and aluminium belong to this eroun* uT- a”'' ‘^**"‘’** metals, alkaline earths

Electrostatic forces predominate in cn acceptor properties,

mall ions of high charge are inrticuk"r formation, so interactions between

TJ)usfluoro-complc,xesLparli‘cuhrt and lead to stable comple.xcs.

• mmonia which has a .smaller di'onN^ * more strongly bound than

lendcncy to form complexes since tlfcv nT'’"’^'"'’ cyanide ions have little

* nnm compete successfully with h wiV " ' nli^aline .solutions where thev

2- Cations with completely r, ''i

|mpper(I), si|ycr(l) and gold(I) which ^^fo'-'’i'v'lls. Typical of this group arc
ns have high polarising power amt o' ”k'* ^ acceptor properties. The.se

appreciable covalent character Cnn formed in their complexes have

metal and the less clccfrl ' " ■

ncicssclectroncgatiyeihedonorr

258

r atom of the ligand; thus cadmium(ll)

TITRIMETRIC ANALYSIS X, 21

and mercury(II) form strong complexes with I and CN ions, but weak
complexes with F".

3. Transition metal ions with incomplete d sub-shells. In this group both Class
A and Class B tendencies can be distinguished. The elements with Class B
characteristics form a roughly triangular group within the Periodic Table, with
the apex at copper and the base extending from rhenium to bismuth. To the left of
this groupj elements in their higher oxidation states tend to exhibit Class A
properties, while to the right of the group, the higher oxidation states of a given
element have a greater Class B character. ' •

On this basis, it is seen that Class A acceptors prefer to bind hard bases, e.g.,
with nitrogen, oxygen and fluorine donor atons, whilst Class B acceptors prefer to
bind to the softer bases, e.g., P, As, S, Se, Cl, Br, I donor atoms. Examination of the
Class A acceptors shows them to have the following distinguishing features ; small
size, high positive oxidation state, and the absence of outer electrons which are
easily excited to higher states. These are all factors which lead to low
polarisability, and such acceptors are called hard acids. Class B - acceptors,
however, have one or more of the following properties: low positive or' zero
oxidation state, large size, and several easily excited outer electrons (for metals
these are the d electrons). These are all factors which lead to high polarisability,
and Class B acids may be called soft acids. , - '

(b) Characteristics of the ligand Among the characteristics of the ligand
which are generally recognised as influencing the stability of complexes in which
it is involved are (i) the basic strength of the ligand, (ii) its chelating properties (if
any), and (iii) steric effects. From the point of view of the analytical applications of
complexes, the chelating effect is of paramount importance and therefore merits
particular attention. > ' . ■

The term ‘chelate effect’ refers to the fact that a chelated complex, i.e., one
formed by a bidentate or a multidentate ligand, is more stable than the
corresponding complex with monodentate ligands: the greater the number of
points of attachment of ligand to the metal ion, the greater the stability of the
complex. Thus the complexes formed by the nickel(II) ion with (a) the
monodentate NHj molecule, (6) the bidentate ethylenediamine (i,2-diaminoe-
thane), and (c) the hexadentate ligand ‘penten’ {(H 2 N-CH 2 -CHi) 2 N-CH 2 -CH 2 -N
(CH 2 'CH 2 -NH 2 ) 2 } show an overall stability constant value for the ammonia
complex of 3.1 x 10®, which is increased by a factor of about 10^° for the complex
ofhgand (h), and is approximately ten times greater still for the third complex

259

X, 22 QUANTITATIVK INORGANIC ANALYSIS

The chelate cfTcct can often be allributed to the increase in entropy which
accompanies chelation; in this context the cJispiaccmcnt of water molecules from
the hydralcri ion must be Ijornc in mind.

The most common stcric effect i.s that of inhibition of comple.x formation owing
to llic presence of a large group cither attached to. or in close proximity to. the
donor atom.

A further factor which must also l>e tiikcn into consideration from the point of
view of the analytical applications of complexes and of complex-formation
reactions is the rate of reaction : to be analytically useful it is usually required that
the reaction be nipid. An important cla.ssificalion of complexes is based upon the
rate at which they undergo substitution reactions and leads to the two groups of
labile and inert complexes. The term labile complex i.s applied to lho.se cases where
nucleophilic substitution is complete within the time required for mixing the
reagents. Thus, for example, svhen e.xcc..ss of aqueous ammonia is added to an
aqueous solution of copper(ll) sulphate the change in colour from pale to deep
blue is instantaneous: llic rapid rcplaa*mcnt of water molecules by ammonia
indicates that the Cu(I!) ion forms kincticnlly labile compic.xcs. Tlic term inert is
applied to those comple.vcs which undergo slow substitution reactions, Lc.,
reactions with half-times of the order of hours or even days at room tcmiwraturc.
Thus the Crflll) ion forms kinctically inert complexes, .so that the replacement of
water molecules coordinated to Ciilll) by other lig.ituls. is a very .slow process at
room temperature.

Kinetic inertness or lability is innucnced by many factors, but the following
general observations form a convenient guide to the bch;iviour of the complexes
of various elements.

(i) Main group elements visually form labile complexes.

(ii) With the exception of Cr(lll)and Codll). most (irst-row transition elements
form labile complexes.

(iii) Second- and third-row transition elements lend to form inert complexes.

For a full discussion of itic topics introduced in this Section a textbook of

Inorganic Chemistry (c.g.. Kef 4), or one dealing, witli complexes (c.g.. Ref 5).
should be consulted.

X, 22. A SIMPLE COMPLEXATION TITRATION. A simple e.xamplc of
the application ofacomplexation reaction to a titration procedure is the titration
of cyanide with silver nitrate solutions, a method first propo.sed by Liebig. When
a solution of .sih^er nitrate is added to a solution containing cyanide ions (c.g., an
alkali cyanide) a white precipitate is formed when the two liquids first come into
contact with one another, but on stirring it re-dissolves owing to the formation of
a stable complex cyanide, the alkali salt of which is soluble;

Ag" +2CN~ =i(Ag(CN)j]~

\yhcn the .above reaction is complete, further addition of silver nitrate solution
yields the insoluble silver cyanoargentalc (sometimes termed insoluble silver
cyanide): thecnd-poinl of the reaction is therefore indiaited bv t)ic formation ofa
permanent precipitate or turbidity.

The only dinicuUy in obtaining a .sharp cnd-poinl lies in the fact that silver
cyanide, precipilaicd by local excess concentration of silver ion somewhat prior
to the equivalence point, is very slow to rcdissolvc and the titration is time-
consuming. In the Deniges modification, iodide ion (usually ns Kl, ca. 0,01 jM) is

TITRIMETRIC ANALYSIS X, 23

used as the indicator and aqueous ammonia (ca. 0.2M) is introduced to dissolve

the silver cyanide. . • . .

The iodide ion and ammonia solution are added before the titration is
commenced; the formation of silver iodide (as a turbidity) will indicate the end
point;

[Ag(NH3)2]++r ^AgI + 2NH3 . . :

During the titration any silver iodide which would tend to form will be kept in
solution by the excess of cyanide ion always present until the equivalence point is
reached :

AgI + 2CN- ^[Ag(CN)2]" fl'

The method may also be applied to the analysis of silver halides by dissolution
in excess of cyanide solution and back-titration with standard silver nitrate. It can
also be utilised indirectly for the determination of several metals, notably nickel,
cobalt, and zinc, which form stable stoichiometric complexes with cyanide ion.
Thus if a Ni(II) salt in ammoniacal solution is heated with excess of cyanide ion,
the )Ni(CN) 4 )^“ ion is formed quantitatively; since it is more stable than the
[Ag(CN) 2 ]~ ion, the excess of cyanide may be determined by the Liebig-Deniges
method. The metal ion determinations are, however, more conveniently made by
titration with EDTA : see the following Sections.

X, 23. COMPLEXONES. The formation of a single complex species rather
than the stepwise production of such species will clearly simplify complexometric
titrations and facilitate the detection of end-points. Schwarzenbach (Ref. 6)
realised that the acetate ion is able to form acetato complexes of low stability with
nearly all polyvalent cations, and that if this property could be reinforced by the
chelate effect, then much stronger complexes would be formed by most metal
cations. He found that the aminopolycarboxylic acids are excellent complexing
agents; the most important of these is 1,2-diaminoethanetetra-acetic acid
(ethylenediaminetetra-acetic acid). The formula (I) is preferred to (II), since it has
been shown from

HOOC-CH,

"OOC— CH

HOOC— CH,

H— N— CHj— CH,— N— H
/ \

CH,— COO‘

(I)

CHz— COOH
CH,— COOH

CH,— COOH

(II)

measurements of the dissociation constants that- two hydrogen atoms are
probably held in the form of zwitterions. The values of pR are respectively pK,
P ^2 = 2.7, = 6.2, and pK^. = 10.3 at 20 °C; these values suggest that

It behaves as a dicarboxylic acid with two strongly acidic groups and that there
are two ammonium protons of which the first ionises in the pH region of about 6.3
and the second at a pH of about 1 1.5. Various trivial names (see Ref. 19) are'used

261

X, 23 QUANTITATIVE INORGANIC ANALYSIS

for dhylencf/ia!nincfclni-<icclic acid and its sodium salts, and these include Trilon
B, Coriiplcxone III, Scqucstrcne, Vcrsenc, and Chclaton 3; the disodium salt is
most widely employed in titrimelric analysis. To avoid the con.stant u.se of the
long name, the abbreviation EDTA is utilised for the disodium salt.

Other complc.xing agents (complexone.s) which arc sometimes used include (a)
nitrilotriacctic acid (HI.) (NIT.A or NTA or Complexone I; this has pA', = 1,9,
pA'j - 2.5, and pA'j ~ 9.7), (/>) tran,s-l,2-diaminocyclohcxanc-A’.N,i\’’,iV',-ietra-
acelic acid (IV): this should presumably be formulated as a zwittcrion structure
like (1); the tibbreviated name is CDTA, DCyTA or DCT A or Complexone IV).

C!I,-COO!I
./ ‘

It— N“Cn,~coo'
\ '

Cl Ij— coon
(in’)

CIIj-COOH

CIIj N
/ \ / \

H.C CH CII.-COOII

1 1

HjC cii cn.— coon

\ / \ / ‘

CII. N

Cll.-COOIt

(IVl

(e) 2,2'-ethylencdioxybis{eihyliminodi(acctic acid)) (V) akso known as ethylene
glycolbis(2-aminocthy! cther)N,N.N’.N'-tetrn-acctic acid (MGTA), and (J)
triethy!cnctctraminc-N.N,N'.N~,N"',N’’'-hcxa-accticacid (TTHA). (VI).

Cl t, COO
■

IIN

/ \

/ CII, coon

(cn.),

(CH,),

(CH,},

C!I,COO)l

CM, coo-
(V) EGTA

CILCOO'

IIN

/ '''

/ CII. coon
(cn.i,

\ " Cl I, cot r

\./ ‘

(VH.):

\ cn.coo

IIN

(CH.(,

CII, coon

HN^

cn.coo '
(VII rruA

CD! A often forms stronger metal complexes than docs EDTA and thus finds
applications in analy.sis, but the metal complexes arc formed rather more slowly
than with EDTA so that the end point of the titration tends to be drawn out with
the former reagent. EGTA finds analytical application mainly in the
etcrmination of calcium in a iTii.xturc of calcium and magnesium and is probably
■superior to EDTA in the calcium/magnesium watcr-hardnc-ss titration (Section
X, 6,3). TTHA forms 1 : 2 complexc,s with many trivalent cations and with some

262

TITRIMETRIC ANALYSIS X, 23

divalent metals, and Pfibil and Vesely (Ref.
7 ) have devised procedures for determining
the components of mixtures of certain ions
without the use of masking agents (see
Section X, 27).

However, EDTA has the widest general
application in analysis because of its
powerful complexing action and com-
mercial availability. The spatial structure
of its anion, which has six donor atoms,
enables it to satisfy the coordination
number of six frequently encountered
among the metal ions and to form strain-
less five-membered rings on chelation. The
resulting complexes have similar structures, but differ from one another. in the
charge they carry. One such structure suggested for the complex with a divalent
ion is shown in Fig. X, 8 ; this structure shows the complex ion exhibiting the
maximum chelating power as a hexa-dentate ligand, but this may not be true for
all metal-EDTA complexes.

To simplify the following discussion EDTA is assigned the formula H 4 Y : the

disodium salt is therefore NajHjY and affords the complex-forming ion HjY^"
in aqueous solution; it reacts with all metals; in a 1:1 ratio. The reactions with
cations, e.g., , may be written as :

+ + Ha Y^ - ^ MY^ " 2 H +

( 1 )

For other cations, the reactions may be expressed as:

■

+ + Ha Y^ “ M Y ■ + 2 H +

( 2 )

+H 2 Y^“ ^MY-f 2H+

(3)

or M'’+-fH 2 Y^" ^(MY)<"'‘‘>++ 2 H+

(4)

One mole of the complex-forming H 2 Y^~ reacts in all cases with one mole of the
metal ion and in each case, also, two moles of hydrogen ion. are. formed. It is
apparent from equation (4) that the dissociation of the complex will be governed
by the pH of the solution ; lowering the pH will decrease the stability of the metal-
EDTA complex. The more stable the complex, the lower the pH at which an
EDTA titration of the metal ion in question may be carried out. Table X, 7
indicates minimum pH values for the existence, of EDTA complexes of some
selected metals.

1_ CO

Fig.X,8

Table X, 7. Stabili ty with respect to pH of some metal-EDTA complexes

Minimuni pH at which Selected metals
complexes exist

1-3

4^'6

8-10

Zr* ; Hi* + ; Th'‘ ; BF + ; Fe^

Pb^ + ; Cu^ ■" ; Zn^ + ; Co^ + ; Ni^ + ; Mn^ * ; Fe^ ; A|3 + • Cd^ + ■ Sn^ +

Ca^+;Sr^+;Ba^-^;Mg^+

263

X, 24 QUANTITATIVE INORGANIC ANALYSIS

I( is thus seen that, in general, EDTA complexes with divalent metal ions are
stable in alkaline or slightly acidic solution, whilst complexes with tri- and letra-
valent metal ions may exist in solutions of much higher acidity,

X. 24. STAniUrV CONSTANTS OF FDTA COMFLFXF^. The stability
of a complex is diaraclcriscd by the stability constant {or formation constant) K:

M"" +Y-*'- ei (S)

K (6)

Some values for the stability constants (expressed as lopK) of mctal-EDTA
complexes are collected in Tabic X. X: these apply to a medium of ionic .strength 1
^ 0.1 at 20 "C,

Table X, 8. .StabiJit) constants ofmclal KOTA complexes

Me-’*

K.7

Zn*' •

!f<7

157

Ca* '

10.7

C<!=-

If. 6

t.ii’-

20 0

Sr=‘

S.f.

Hr*

;i.‘)

Sc"

2J1

fla*'

Vh''

ISO

c:.x’*

205

Mn--

Al’*

If.?

In’*

re* *

14 .t

Pc’*

251

77i**

2yi

Co'*

I6..1

yj-

IS2

7.3

NV*

IK 6

Ct'*

24 0

i.i*

Cu"

IS.S

<Y’'

150

N;i*

1.7

In equation (6) only the fully ionised form of EDTA. i.c.. theion Y"*'. has been
taken into account, but at low pH values the species HV''". HjV" and

even undissocialcd H^Y may well be present; in other words, only a part of the
EDTA uneombined with metal may be present as Y^'. Further, in equation (6)
the metal ion M"‘ is assumed to be uncomplexcd. i.e., in aqueous solution it is
simply present as the hydrated ion. If. however, the solution also contains
substances other ihan EDTA which can complex with the metal ion, then the
whole of this ion uneombined wjih EDT.A may no longer be present as the simple
hydrated ion. Thus, in pracuce, the stability of metabEDTA complexes may be
altered by (u) variation in pi I and {b) by the presence of other complexing agents.
The stability constant of tlie EDTA complex \t ill tlicn be difierent from the value
recorded for a specified pH in pure aqueous solution; the value recorded for the
new conditions is termed the apparent or conditional sfabnity constant. It is clearly
neccs.sary to examine the effect of these two factors in some detail.

(a) pH effect. Tire apparent stability constant at a given pH may be
calculated from the ratio K/x, where a is the ratio of the total uneombined EDTA
(in all forms) to the form Y'* . llius K,,, the apparent stability constant for the
mclal-EDTA complex at a given pH. can be calculated from the expression

log f^ii = log X - loga (7)

TTic factor a can be calculated from the known dissociation constants of EDTA.
and since the proportions of the variou,s ionic species derived from EDTA will be
dependent upon the pH of the solution, a will also varv with pH: a plot of ioga
against pH shows a variation of loga = 18 at pH - I'to loga == 0 at pH = 12;
such a curve is very useful for dealing with calculations of apparent stability

264

TITRIMETRIC ANALYSIS X, 25

constants. Thus, for example, from Table X, 8 log K of the EDTA complex of the

ion is 18.0 and from a graph of log a against pH, it is found that at a pH
= 5.0, log a = 7. Hence from equation (7), at a pH of 5.0 the lead-EDTA complex
has an apparent dissociation constant given by : '

logXH = 18.0-7.0= 11.0.

Carrying out a similar calculation for the EDTA complex of the Mg^ ion (log K
= 8.7), for the same pH (5.0), it is found;

log XH(Mg(II)-EDTA) = 8.7 - 7.0 = 1 .7.

These results imply that at the specified pH the magnesium complex is
appreciably dissociated, whereas the lead complex is stable, and clearly titration
of an Mg(II) solution with EDTA at this pH will be unsatisfactory, but titration of
the lead solution under the same conditions will be quite feasible. In practice, for a
metal ion to be titrated with EDTA at a stipulated pH the value of log should
be greater than 8 when a metallochromic indicator is used.

As indicated by the data quoted in the previous Section, the value of log a is
small at high pH values, and it therefore follows that the larger values of log
are found with increasing pH. However, by increasing the pH of the solution the
tendency to form slightly soluble metallic hydroxides is enhanced owing to the
reaction;

(MY)(«-4)+ ^„OH“ ^M(OH)„ + Y^-..

The extent of hydrolysis of (MY)'"”'^*'*' depends upon the characteristics of the
metal ion, and is largely controlled by the solubility product of the metallic
hydroxide and, of course, the stability constant of the complex.. Thus iron(III) is
precipitated as hydroxide {K^oi = 1 x 10"^®) in basic solution, but nickel(II), for
which the relevant solubility product is 6.5 x 10~ remains complexed. Clearly
the use of excess EDTA will tend to reduce the effect of hydrolysis in basic
solutions. It follows that for each metal ion there exists an optimum pH which will
give rise to a maximum value for the apparent stability constant.

(b) The effect of other complexing agents. If another complexing agent (say
NHj) is also present in the solution, then in equation (6) [M"'*'] will be reduced
owing to complexation of the metal ions with ammonia molecules. It is
convenient to indicate this reduction in effective concentration by introducing a
factor p, defined as the ratio of the sum of the concentrations of all forms of the
metal ion not complexed with EDTA to , the concentration of the simple
(hydrated) ion. The apparent stability constant of the metal-EDTA complex,
taking into account the effects of both pH and the presence of other complexing
agents, is then given by; -

logX^z = logX-loga-log/J. (8)

X, 25. TITRATION CURVES. If, in the titration of a strong acid, pH is
plotted against the volume of the solution of the strong base added, a point of
inflexion occurs at the equivalence point (compare Section X, 12). Similarly, in the
til’^ation, if pM (negative logarithm of the ‘free’ metal ion concentration;
pM = -log[M"+]) is plotted against the volume of EDTA solution added, a
point of inflexion occurs at the equivalence point; in some instances this sudden
increase may exceed 10 pM units. The general shape of titration curves obtained

265

X, 26 QUANTJTATrvn INORGANIC ANALYSIS

by titrating 1 0.0 cm^ of a 0.0 1 .\f solution
of a metal ion M with a O.Olilf-EDTA
.solution is .shown in Fig X. 9. Tlie
apparent stability conslant.s of various
mcial-EDTA comple.xcs arc indicated
at the e-xiremc riplit of the curves. It is
evident that the greater the siabilitv
constant, tlic .sharper is the end-point
provided the pH is maintained constant.

In acid-base titrations the end-point is
genendly detected bv a pH-scnsiiivc
indicator. In the CDTA titration a metal
!on-scnsiiivc indicator (abbreviated to
fiwuil imliccitor or metal ion indicaitor) is
often employed to detect changes of
pM. .Such indicators (which contain
l>pes of chelate groupings and generally
Kip X 9 possc.ss resonance .sy.sterns tvpical of

menUon.: r„.„

Tl.; c,rpZ cf ,i,

delcrmincj is^biiff're'ho UiJ.l'c nT liic melul ion to be

»nd titrnted dtr^ •'t"**-'!' f'”-'

prevent prccininiionoftiii>i,v i ■ i }'} •‘'‘^lution. It m;iy be necessary to
of .some auxilia'rv compicsinc [hemctal(ora basic.salt)by iheaddition

amine. At tlm equi^""ele Z / V Iricthanol-

ion being determined decreases

changcincolourofamctalindu-wer ^ determined by the

point may also be determined hv * " * ‘’osponds to changes in pM : the end

mettk. or, in »™e core., by pmcmZZ7S,St'‘''’"'‘'''‘'

dirccilv; thus thev n°aj' prednii-!y!f^*^ cannot, for various reasons, be titrated
for the titration. S «l>c solution in the pH range neces.sary

not available. In such'ca.s'c.s an metal indiealor is

resulting solution is bufl'ered* 1 1 ii FDTA solution is added, the

sulphate or of mamiCM'um dilnri t .* a .solution of zinc chloride or

end-point is delected with the aid^'of the Purpose. The

metal ion introduced in the back-iitraiion ' '‘csponds to the

used for metal ionMhai'do'’nm tilrations may be
indicator, or for metal inne ».?, i ^ unsatisfactorily) with a metal

than those of other metals suclTas comple.xcs that arc more stable

to be determined mav L. ‘ - ^pncsium and calcium. The metal cation
■emuned may be treated with the magnesium complex of EDTA,

266

TITRIMETRIC ANALYSIS X, 27

when the following reaction occurs:

M"+ +Mg^+

The amount of magnesium ion set free is equivalent to the cation present and can
be titrated with a standard solution of EDTA and a suitable metal indicator.

An interesting application is the titration of calcium. In the direct titration of
calcium ions Solochrome Black (Eriochrome Black T) gives a poor end-point; if
magnesium is present, it is displaced from its EDTA complex by calcium and an
improved end-point results (compare Section X, 55).

D. Alkalimetric titration. When a solution of disddium ethylenediamine-
tetra-acetate, Na2H2 Y, is added to a solution containing metallic ions, complexes
are formed with the liberation of two equivalents of hydrogen ion:

-f ^(MY)'"-^>+ -t-2H+

The hydrogen ions thus set free can be titrated with a standard solution of
sodium hydroxide using an acid-base indicator or a potentiometric end-point;
alternatively, an iodate-iodide mixture is added as well as the EDTA solution
and the liberated iodine is titrated with a standard thiosulphate solution.

The solution of the metal to be determined must be accurately neutralised
before titration; this is often a difficult matter on account of the hydrolysis of
many salts, and constitutes a weak feature of alkalimetric titration.

E. Miscellaneous Methods. Exchange reactions between the tetracyano-
nickelate(II) ion [Ni(CN)4]^~ (the potassium salt is readily prepared) and the
element to be determined, whereby nickel ions are set free, have a limited
application. Thus silver and gold, which themselves cannot be titrated
complexometrically, can be determined in this way.

[Ni(CN)4]^- +2Ag+ ^2[Ag(CN)2r +Ni"+

These reactions take place with sparingly soluble silver salts, and hence provide a
method for the determination of the halide ions Cl“, Br~, I~, and the thiocyanate
ion SCN~. The anion is first precipitated as the silver salt, the latter dissolved in a
solution of [Ni(CN)4]^ ", and the equivalent amount of nickel thereby set free is
determined by rapid titration with EDTA using an appropriate indicator
(Murexide, Bromopyrogallol Red).

Sulphate may be determined by precipitation as barium sulphate or as lead
sulphate, the precipitate is dissolved in an excess of standard EDTA solution, and
the excess of EDTA is back-titrated with a standard magnesium or zinc solution
using Solochrome Black (Eriochrome Black T) as indicator.

Phosphate may be determined by precipitating as Mg(NH4)P04,6H20,
dissolving the precipitate in dilute hydrochloric acid, adding an excess of
standard EDTA solution, buffering at pH = 10, and back-titrating with standard
magnesium ion solution in the presence of Solochrome Black.

X, 27. TITRATION OF MIXTURES, SELECTIVITY, MASKING AND
DEMASKING AGENTS. EDTA is a very unselective reagent because it
complexes with numerous di-, tri-, and tetra-valent cations. When a solution
containing two cations which complex with EDTA is titrated without the
addition of a complex-forming indicator, and if a titration error of 0.1 per cent is
permissible, then the ratio of the stability constants of the EDTA complexes of the
two metals M and N must be such that ^ 10® if N is not to interfere with

267

X, 27 QUANTITATIVE INORGANIC ANALYSIS

the titration of M. Strictly, of course, the constants K^, and considered in the
above expression should be the apparent stability constants of the complexes. If
complex -forming indicators arc used, then for a similar titration error K,,,,/Kv

g JO’’-

The following procedures will help to increase the .selectivity:
fa) Ry .suitable control of the pH of the solution. This, of cour.se. makes use
of the different .siabilitic.s of metal -I-DTA conipic.xcs. Thus bismuth and thorium
can be titrated in an acidic solution (pH == 21 with Xylcnoi Orange or
Mclluithymol Blue as indicator and most divalent cations do not interfere. A
mixture of bismuth and lead ions can be successfully titrated by first titrating the
bismuth at pH 2 witli Xylcnoi Orange as indicator, and then adding hexamine to
raise the pH to about 5. and titrating the lead (.see Section X. 72),

(hi By the use of masking agents. Masking may be defined as the proccs.s in
which a substance, witliout pltysica! separation of it or its reaction product.s. is so
tran.sformcd that it doe.s not enter into a particular reaction. Dem.isking is the
process in which the masked substance regains its ability to enter into a particular
reaction.

By the u.sc of masking agents, .some of tlic cations' in a mixture can often be
‘masked’ so that they can no longer react with MD'f'A or with the indicator. An
effective masking agent is the cyanide ion; this forms .stable cyanide complexes
with the ciiion.s of Cd, Zn. Heiil). Cu, Co. NT. .Ag, and the platinum mciai,s. but
not with the alkaline earths, manganese, am! Ic.id;

-t-4CN- ~.(M!CNT,i-'

It is therefore possible to rictcrmine cations such ns Ca* and
Mn^ * in the presence of the above-mentioned metals by masking with an c.xccss
of potassium or sodium cyanide. A small amount of iron may be tn.iskcd by
cyanide if it is first reduced to the ironfll) state by llic addition of a.scorbic acid.
Tiianium(lV). iron(III). and aluminium can be nnisked with irieihanolaminc:
mercury with iodide ions; and aluminium, iron! Ill), litaniutndVj.and tinflll with
ammonium (luondedhc cations ofthenikahne-earth metals yield slightly .soluble
Iluoridcs).

Sometimes the metal may be transformed into a different oxidation state: thus
copperd!) may be reduced in acid solution by hydroxylnmine or ascorbic acid.
After rendering ammoniacal. nickel or cobalt can be titrated using, for example,
mure.xidc as indicator without interference from the copper, which is now present
asCu(l). lron(Ill)can often be similarly masked by reduction w ith ascorbic acid.

(c) The cyanide complexes of zinc and cadmium may be demasked with
formaldchydc-acetic acid solution or, better, witli chloral hydrate;

(Zn(CN)i]-~-f4H* -vdllCHO — .Zn'’* 1 - 4110 CHyCN

Tlic use of masking and selective demasking agents permits the successive
titration of many metal.s. Thus a solution containing Mg. Zn. and Cu can be
titrated as follows:

O’) Add excc.ss of standard I- DT.A and back-tit rale with standard Mg solution
using Solochromc Black (liriochrome Black T) as indicator. This gives the sum of
all the metals present.

(ii) Treat an aliquot portion with excess of KCN and titrate as before. This
gives Mg only,

(iii) Add excess of chloral hydrate (or of formaldehyde-acetic acid solution.

268

TITRIMETRIC ANALYSIS X, 28

3 : 1) to the titrated solution in order to liberate the Zn from the cyanide complex,
and titrate until the indicator turns blue. This gives the Zn only. The Cu content
may then be found by difference.

(d) , Classical separations may be applied if these are not tedious; thus the
following precipitates may be used for separations in which, after being
redissolved, the cations can be determined complexometrically ; CaC 204 ,
nickel dimethylglyoximate, Mg(NH 4 )P 04 , 6 H 20 , and CuSCN.

(e) Solvent extraction is occasionally of value. Thus zinc can be separated
from copper and lead by adding excess of ammonium thiocyanate solution and
extracting the resulting zinc thiocyanate with 4-methyl pentan-2-one (isobutyl
methyl ketone); the extract is diluted with water and the zinc content determined
with EDTA solution.

(f) The indicator chosen should be one for which the formation of the metal-
indicator complex is sufficiently rapid to permit establishment of the end-point
without undue waiting, and should preferably be reversible.

(g) Anions, such as orthophosphate, which can interfere in complexometric
titrations may be removed using ion exchange resins. For the use of ion exchange
resins in the separation of cations and their subsequent EDTA titration, see
Chapter VII.

(h) ‘Kinetic masking’ is a special case in which a metal ion does not
effectively enter into the complexation reaction because of its kinetic inertness
(see Section X, 21). Thus the slow reaction of chromium(III) with EDTA makes it
possible to titrate other metal ions which react rapidly, without interference,
from Cr(III); this is illustrated by the determination of iron(III) and
chromium(III) in a mixture (Section X, 68).

X, 28, METAL ION INDICATORS. General properties. The success of ah
EDTA titration depends upon the precise determination of the end-point. The
most common procedure utilises metal ion indicators. The requisites of a metal
ion indicator for use in the visual detection of end-points include;

(a) The colour reaction must be such that before the end-point, when nearly all
the metal ion is complexed with EDTA, the solution is strongly coloured.

{b) The colour reaction should be specific or at least selective.

(c) The metal indicator-complex must possess sufficient stability, otherwise,
because of dissociation, a sharp colour change is not obtained. -The metal
indicator-complex must, however, be less stable than the metal-EDTA
complex to ensure that, at the end-point, EDTA removes metal ions from the
metal indicator-complex. The change in equilibrium from the metal
indicator-complex to the metal-EDTA complex should be sharp and rapid.

(d) The colour contrast between the free indicator and the metal indicator-
complex should be such as to be readily observed.

(e) The indicator must be very sensitive to metal ions (i.e., to pM) so that the
colour change occurs as near to the equivalence point as possible.

if) "^e above requirements must be fulfilled within the pH range at which the
titration is performed.

Dyestuffs which form complexes with specific metal cations can serve as
indicators of pM values; 1; 1-complexes (metal: dyestuff =1:1) are common, but
2; 1-complexes also occur. The metal ion indicators, like
LDTA Itself, are chelating agents; this implies that the dyestuff molecule
possesses several ligand atoms suitably disposed for coordination with a metal

269

X, 28 QUANTITATIVE INORGANIC ANALYSIS

atom. They can, of course, equally take up protons, which also produces a colour
chance: metal ion indicators arc thcrefoTC not only pM but also pH indicators.

Theory of the sisual use of metal ion indicators. Our discussion will be
confined to the more common 1 : 1 -complexes. Tiic use of a metal ion indicator in
an EDTA titration may l>c written as:

M-ln-hEDTA — M-ED7A-Hn

This reaction will proceed if the metal indicator-complex M~In is less stable than
the mctal-EDTA complex M-EDTA. The former dissociates to a limited extent,
and during the titration the free metal ions are progressively compicxcd by the
EDTA until ultimately the metal is displaced from the complex M-Tn to leave the
free indicator (In). T})c stability of the metal indicator-complex may be c.xpressed
in terms of ibcformation consiaiu (or iiuliriuori onstaiu)

- lM- ln],'lM)[ln)

The indicator colour change is affected by tlie hydrogen-ion concentration of
the .solution, and no account ofthis has been taken in the above expression for the
formation constant. Tlius Solochromc Black fEriochromc Black T). which may
be written as H-ln ‘ .exhibits the following acid-base behaviour:

* 5 3 1 .}

Reel niuc

j -n _
"to s‘-T;,5

In^'

Yellow -oranf;'

In the pH range 7-1 1, in wliich the dye itself exhibits a blue colour, many metal
ions form red complexes; these colours arc extremely sensitive, ns is shown, for
example, by the fact that molar solutions of magnesium ion give a

distinct red colour with the indic;itor. From the practical viewpoint, it is more
convenient to define the appurfiit imlicaior consumt K’l^, whicii varies with pH,
as:

X',„ = (MIn-)‘[M'’'‘JiInj

where {Min ] concentration of metal ion-indicator complex,

[M"* J ~ concentration of metallic ion, and
[In] = conecntraiion ofindicator not compicxed with metallic ion.

(This, for the above indicator, is cqu:il to [H -In “ ) -f [H 1 n= ■■] + [In* ~ ).)

The equation may be expressed .as;

lof: A"',., = pM +log(MIn~HInl;

A gives the value of pM when half the lota! indicator is prc.sent as the metal
ion complex. Some values for log K\^ for Cain" and Mgln” rtaspectivcly (where
H 2 ln is the anion of Solochromc Black (Eriochrornc Black T)arc: 0.8 and 2.4 at
pH = 7; 1.9 and 3.4 at pH .S; 2,8 and 4.4 at pH =. 9; 3.8 and 5.4 at pH = 10: 4.7
and 6,3 at pH = Il;5.3and6.8atpH — 12. For a small titration error K'i„ should
he large ( > 10 ), the ratio of the app.ircnt stability constant of the nietal-EDTA
complex to that of the metal indicator-complex A'’i„ should be large (> 10^).
and the ratio of the indicator concentration to the metal ion concentration should
be small {<10“^).

The visual mctallochromic indicators discu.sscd above form by far the most
important group of indicators for EDTA titrations and the operations

270

TITRIMETRIC ANALYSIS X, 28

subsequently described will be confined to the use of indicators of this type;
nevertheless there are certain other substances which can be used as indicators
(see Ref. 8). - - .

Some examples of metal ion indicators. Numerous compounds have been
proposed for use as pM indicators; a selected few of these will be described.
Where applicable. Colour Index (C.l.) references are given (Ref. 9). It has been
pointed out by West (Ref. 8), that apart from a few miscellaneous compounds, the
important visual metallochromic indicators fall into three main groups: (a)
hydroxyazo compounds, (b) phenolic compounds and hydroxy-substituted
triphenylmethane compounds, (c) compounds containing an aminomethyldi-
carboxymethyl group: many of these are also triphenylmethane compounds.

Note. In view of the varying stability of solutions of these indicators, and the
possible variation in sharpness of the end-point with the age of the solution, it is
generally advisable (if the stability of the indicator solution is suspect), to dilute
the solid indicator with 100-200 parts of A.R. potassium (or sodium) chloride,
nitrate or sulphate (potassium nitrate is usually preferred) and grind the mixture
well in a glass mortar. The resultant mixture is usually stable indefinitely if kept
dry and in a tightly stoppered bottle.

Murexide, (C.l. 56085) This is the ammonium salt of purpuric acid, and
its anion has the structure (I). It is of interest because it was probably the first
metal ion indicator to be employed in the EDTA titration. Murexide solutions
are reddish-violet up to pH = 9 (H4D“), violet from pH 9 to pH, 11 (H3D^“),and
blue-violet (or blue) above pH 11 (HjD^"). These colour changes are probably
due to the progressive displacement of protons from the imido groups ; since there
are four such groups, murexide may be represented as H4D“. Only two of these
four acidic hydrogens can be removed by adding an alkali hydroxide, so that only
two pK values need be considered; these are pK^ = 9.2 (HjD" — > HsD^") and
pKj = 10.5 (HjD^ " — » H2D^ “ ). The anion H4D " can also take up a proton to
yield the yellow and unstable purpuric acid, but this requires a pH of about 0.

0-0

HN— C C— NH

/ % / \

0=c C— N=C C=0

\ / \ /

HN— C C— NH

^ ^ .
o o

(I)

Murexide forms complexes with many metal ions: only those with Cu, Ni, Co,
Ca and thelanthanoids are suificiently stable to find application in analysis. Their
colours in alkaline solution are orange (copper), yellow (nickel and cobalt), and
red (calcium); the colours vary somewhat with the pH of the solution.

Murexide may be employed for the direct EDTA titration of calcium at pH
= 11; the colour change at the end point is from red to blue- violet, but is far from
ideal. The colour change in the direct titration of nickel at pH 10-11 is from
yellow to blue- violet.

Aqueous solutions of murexide are unstable and must be prepared each day.
The indicator solution may be prepared by suspending 0.5 g of the powdered

thoroughly, and allowing the undissolved portion tp
settle. The saturated supernatant liquid is used for titrations. Every day the old

271

X, 2S QUANTITATIVE INORGANIC ANALYSIS

atom. They can. of course, equally take up protons, which .also produces a colour
change; mcla! ion indicators arc therefore not only pM but also pH indicators.

Iliforj' of the \isual use of metal ion indiaitors. Our di.scussion will be
confined to* the more common 1 : 1-compIc.xcs. The u.se of a metal ion indicator in
an EDTA titration may be written as:

M-Jn-fEDTA — * M-hDT.A + It)

This reaction will proceed iftheincial indicator-oomple.x M -In is lcs.s .stable than
the metal -EDTA comple.x M-EDTA. The former dissociates to a limited c.xtent,
and during tlic titration the free metal ions arc progressively complexed by the
EDTA until ultimately tlie metal isdi.splaccd from thccomplc.x M-ln to lease the
free indicator (In). The .stability of the metal indicator -<omp!c.x may lx; expressed
in terms of the faniutlinn mnsfu/n (or indkaiar corneanr} A'j^:

AV. - [NMnl iM]{In!

Tlic indicator colour change is affected by the hydrogen-ion concentration of
the solution, and no account of this has been taken in the above expression for the
formation constant. Thus Solochrorne Black (Eriochromc Black T). which may
be written :is IDln exhibits the following acid -ba.sc behaviour:

Tj 1 - f" tit f’" .

H,In eeerre Hln* ccr-ene:;:
5.1 '.t m5-i:.s

Red liluc

In'-
Yellott -oran;*-

In the pH range 7- 1 1. in which the dye itself exhibits a blue colour, many metal
ions form red complexes; these colours arc extremely sensitive, as is .shown, for
example, by the fact that 10 -lO' ' molar solutions of magnesium ion give a
distinct red colour with the indicator. From the practical viewpoint, it Is more
convenient to define the tipparciu huHfutor constant A",,, which varies with pH,
a.s;

A',„=:(Mln-].iM--]flnl

where [Min | = concentration of rneial ion- indicator complex.

[M"* ] -■= concentration of metallic ion, and
(In) - concentration of indicator not complexed with metallic ion.

(Thi.s, for the above indicator, is equal to [! Mn • j -f (H In' - ] -I- [In-' "J.)

The equation may be expressed a.s:

= pM-r!og[Mln'HIn];

Ai |„ gives the value of pM wlien half the total indicator is present as the metal
ion complex. Some values for log K’,^ for Cain - and Mnln " respectively (where
Hjln ' IS the anion of Solochrorne Black (Eriochromc Black T) arc: 0,8 and 2.4 at
pH = 7; 1.9 and 3.4 .at pH -= 8:2.8 .and 4.4 at pH = 9; 3.8 and 5.4 at pH = 10:4.7
anci6,3atpH == lI;5.3.and6,8.aipH = 12. ForasmalltitrationcrrorA'’iaShould
be large ( > 10 ), the ratio of the apparent stability constant of the mctal-EDTA
complex A' j,,y to that of the metal indicator-complex A'i„ should be large (> 10*1,
and the ratio of the indicator concentration to the metal ion conccntr.ation should
be small (< 10'^).

Tlic visual mctallochromic indicator,s discussed above form by far the most
important group of indicators for EDTA titrations and the operations

270

TITRIMETRIC ANALYSIS X, 28

subsequently described will be confined to the use of indicators of this type;
nevertheless there are certain other substances which can be used as indicators

(see Ref 8). ,

Some examples of metal ion indicators. , Numerous compounds have been
proposed for use as pM indicators; a selected few of these will be described.
Where applicable, Colour Index (C.I.) references are given (Ref 9). It has been
pointed out by West (Ref 8), that apart from a few, miscellaneous compounds, the
important visual metallochromic indicators fall into three main groups: (a)
hydroxyazo compounds, (b) phenolic compounds and hydroxy-substituted
triphenylmethane compounds, (c) compounds containing an aminomethyldi-
carboxymethyl group: many of these are also triphenylmethane compounds.

Murexide. (C.I. 56085) This is the ammonium salt of purpuric acid, and
its anion has the structure (I). It is of interest because it was probably the first
metal ion indicator to be employed in the EDTA titration. Murexide solutions
are reddish-violet up to pH = 9 (H4D “), violet from pH 9 to pH 11 (H3D^“), and
blue-violet (or blue) above pH 11 (HjD^"). These colour changes are probably
due to the progressive displacement of protons from the imido groups; since there
are four such groups, murexide may be represented as H4D“. Only two of these
four acidic hydrogens can be removed by adding an alkali hydroxide, so that only
two pK values need be considered; these are PK4 = 9.2 (H2D" — > HjD^") and
PK3 = 10.5 (H3D^" — > H2D^“). The anion H4D" can also take up a proton to
yield the yellow and unstable purpuric acid, but this requires a pH of about 0.

O' o

HN— C C— NH

0=c C— N=C C=0

\ / \ /

HN— C C— NH

o o

( 1 )

Murexide forms complexes with many metal ions: only those with Cu, Ni, Co,
Ca and thelanthanoids are sufficiently stable to find application in analysis. Their
colours in alkaline solution are orange (copper), yellow (nickel and cobalt), and
red (calcium); the colours vary somewhat with the pH of the solution.

^Murexide may be employed for the direct EDTA titration of calcium at pH
-11; the colour change at the end point is from red to blue- violet, but is far from
ideal. The colour change in the direct titration of nickel at pH 10-11 is from
yellow to blue-violet.

Aqueous solutions of murexide are unstable and must be prepared each day.
The indicator solution may be prepared by suspending 0.5 g of the powdered
yestuff in water, shaking thoroughly, and allowing the undissolved portion to
settle. The saturated supernatant liquid is used for titrations. Every day the pld

271

X, 28 quantitative INORGANIC ANALYSIS

supcrnalant liquid is decanted 'and the residue treated with water as before to
provide a fresh solution of tlic indicator. Alternatively, one may prepare a
mixture of the indicator with pure sodium chloride in the ratio I tffK), and employ
0,2'U4 g in e.ach titration. A screened indicator, consisting of 0.2 g ofmurcxide,
0^5 g of Naphtliol Green B. and lOOg t>f pure sodium chloride ground together to
form a uniformly coloured mixture has been proposed ; about 0.2 g of the mixture
is suitable for 100 cm’ of the sample solution. The colour change for calcium is
from olive-green, through grey, to a sudden blue.

Solachrome Hlach (Kriocliromc Black T). Tliis snb.stance is sodium HI-
hydroxy-2-naphlhy!a7o)-6-nilro-2-naplitho!-4-su!phon3tc(ll): and has the
Cmlour lndcx reference CM. 14645. In strongly acidic solutions the dye tends to
polymerise to a red-hrown product, and consequently tlie indicator is rarely
applied in the HDT.A titration of solutions more acidic than pH ~ 6.5.

/y \

NO;

(111

The sulphonic acid group gives up its proton long before the pH range of 7-12.
which is of immediate interest for rnctal-ion indicator use. Only the dissociation
of tlic two hydrogen atoms of the phenolic groups need therefore be considered,
and so the dyestuff m.ay be represented by the formula H;D ", The two pK values
for these hydrogen atoms are 6.3 and 11,5 resircctivcfy. Below pH = 5.5, the
solution of Solochromc Black {Eriochroine Black T) is red (due to HjD').
between pH 7 and 11 it is blue (due to IID^*). and above pH = 11.5 it is
yellowish-orange (due to D’ ). In the pH range 7 11 thcaddition ofmctallicsalts
produces a brilliant cliangc m colour from blue to red ;

M'* (blue) — Ml)' (rcd)-i- H *

This colour change can be observed with the ions of Mg, Mn. Zn. Cd, Hg. Pb. Cu.
Al, Fc.l i, Co, Ni, and the Pt tnciab lo maintain the pH constant (co. 10) a buffer
mixture is added, and most of tlic above metals must f>e kept in solution with the
aid of a weak complexing reagent such as ammonia or tartrate. The cations of Cu,
Co, Ni, Al. Fcflll), Ti(l V), and certain of the Pi metals form such stable indicator
complexes that the dyestufT can no longer be liberated bv adding EDTA: direct
titration of these ions using Solochromc Black (Frioclirome Black T) as indicator
is therefore impracticable, and the metallic ions are said to ‘block’ the indicator.
However, with Cu. Co, Ni, and Al a back-titration can be carried out. for the rate
of reaction of their EDTA complexes with the indicator is extremely slow and it is
possible to titrate the excess of EDT.A with standard zinc or magnesium ion
solution.

Cu. Ni, Co, Cr, Fc, or Al, even in traces, must be absent when conducting a
direct titration of the other metals listed above; if the metal ion to be titrated does
not react with the cyanide ion or with triethanolamine, these substances can be
used as masking reagents. It has Ijccn stated that the addition of 0.5-1 cm’ of

272

TITRIMETRIC ANALYSIS X, 28

O.OOlM-o-phenanthroline prior to the EDTA titration eliminates the ‘blocking
effect’ of these metals with Solochrome Black (Eriochrome Black T) and also with
Xylenol Orange (see below). '

The indicator solution is prepared by dissolving 0.2 g of the dyestuff in 15 cm^ of
triethanolamine with the addition of 5 cm^ of absolute ethanol to reduce the
viscosity; the reagent is stable for several months. A 0,4 per cent solution of the
pure dyestuff in methanol remains serviceable for at least a month.

It may be noted that the dyestuff in which the hitro group is absent, viz., sodium
l-(l-hydroxy-2-naphthylazo)-2-naphthol-4-sulphonate (Solochrome Black 6B;
Eriochrome Blue-Black B; Colour Index No. 14640) is superior as far as the
solution stability is concerned, and the colour change is sharper with Mg and
certain other metals (Zn and Pb excepted). The colour change is from red to blue;
the indicator may be screened with a little 0.5 per cent aqueous tartrazine
solution, when the resulting end-point is from scarlet (or orange red) to apple
green. A 0.5 per cent ethanolic solution of the dyestuff is stable for at least two
months.

Patton and Reeder’s indicator. The indicator is 2'hydroxy- 1 -(2-hydroxy-4-
sulpho-l-naphthylazo)-3-naphthoic acid (III); the name may be abbreviated to
HHSNNA. Its main use is in the direct titration of calcium, particularly in the
presence of magnesium. A sharp colour change from wine red to pure blue is
obtained when calcium ions are titrated with EDTA at pH values between 12 and
14. Interferences are similar to those observed with Solochrome Black
(Eriochrome Black T), and can be obviated similarly. This indicator may be used
as an alternative to murexide for the determination of calcium.

OH HO COOH

The dyestuff is thoroughly mixed with 100 times its weight of sodium sulphate,
and 1 g of the mixture is used in each titration. The indicator is not very stable in
alkaline solution.

Solochrome Dark Blue or Calcon. (C.1. 1 5705). This is sometimes referred
to as Eriochrome Blue Black RC; it is in fact sodium 1 -(2-hydroxy- 1-
naphthylazo)-2-naphthol-4-sulphonate, as shown in formula (IV); The dyestuff
has two ionisable phenolic hydrogen atoms; the protons ionise stepwise with
pK s of 7.4 and 13.5 respectively. An important application of the indicator is in
the complexometric titration of calcium in the presence of magnesium; this must

OH HO

273

X. 28 QUANTITATIVE INORGANIC ANALYSIS

be carried out at a pH of about 12,3 {obtained, for example, with a dicthylaminc
bufTcr: 5 cmVlfW of solution) in order to avoid the interference of
magnesium. Under these conditions magnesium is precipitated quantitatively as
the hydroxide. The colour change is from pink to pure blue.

The indiailorsahiiloii is prepared by dissolving 0.2 g of the dyestuff in 50cm’
of methanol.

Calrnagite. Tliis indicator, !-{l-hydro,xyl'4-methyl-2-phcnylazo)-2-naph-
tho!-4-sulphonic .acid (V), lias the same colour cliange as .Solochrome Black
(nriochromc Black T), but the colour change is somewhat clearer and sharper. An
import.anl advantage i.s that aqueous .solutions of the indicator are stable almost
indefinitely. It may be substituted for Solochrome Black (Eriochrome Black T)
without change in the experimental proccdure.s for the titration of calcium plus
magnesium (see Sections .X, 55, 64),

Oil Ito

•ll-OjS-H' y-N~N--<;

(Vt

ni,

Calmagitc func!ion,s as an acid--ba.se indicator;

HjD

low pit

Brigtit

red

11, D

tlriptit

red

, pit 7 1

pH ll.-J 13. t

C’Ie.-ir

Wue

RetUnh*

OfJiiKC

The hydrogen of the suiphonicacid group plays no part in the functioning of the
dye as a metal ion indicator. The acid propcnic.s of the hydroxyl groups arc
expressed by pK , ==S.14andpK, =~ 12..i5. 'Die blue colour of Cahmagitc at pH
= 10 i.s changed to red by the addition of magnesium ions, the change being
reversible:

MgD’-

Clcar blue Ucvl

This is the basis of the indicator action in the EDTA titration. The pH = 10 is
attained by the use of an aqueous ammonia-ammonium chloride buffer mixture.

The combining ratio between calcium or macncsium and the indicator is 1:1’.
the magnesium compound is the more stable. Calmagitc is similar to Solochrome
Black in that small amount.s of copper, iron, and aluminium interfere seriously in
the titration of calcium and magnesium, and similar ma.sking agents may be used.
Potassium hydroxide sliould be employed for tiie neutralisation of large amounts
of acid since sodium ions in high concentration cause diflicultv.

The mlicator soluiinii is prepared by dissolving 0,05 g of Ca’imagile in 100 cm^
of water. It is stable for at least 12 months when stored in a polythene bottle out of
sunlight.

Calcichroine. This indicator, cyclotris-T-fl-azo-S-hydroxymaphthalcne-
3,6-disulphonic .acid) (VI), is unusual in having a cyclic structure, and is vcQ'
selective for calcium. It is in fact not very suitable as an indicator for EDT.A

274

TITRIMETRIC ANALYSIS X, 28

titrations because the colour change is not particularly sharp, but if EDTA is
replaced by CDTA (see Section X, 23), then the indicator gives good results for
calcium in the presence of large amounts of barium and small amounts of
strontium (Ref 10).

Fast Sulphon Black F. (C.I. 26990). This dyestuff is the sodium salt of 1-
hydroxy-8-(2-hydroxynaphthylazo)-2-(sulphohaphthylazo)-3,6-disulphonic acid
(VII). The colour reaction seems virtually specific for copper ions. In ammoniacal
solution it forms complexes with only copper and nickel; the presence of
ammonia or pyridine is required for colour formation. In the direct titration of
copper in ammoniacal solution the colour change' at the end-point is from
magenta or (depending upon the concentration of copper(II) ions) pale blue to
bright green. The indicator action with nickel is poor. Metal ions, such as those of
Cd, Pb, Ni, Zn, Ca, and Ba, may be titrated using this indicator by the prior
addition of a reasonable excess of standard copper(ll) solution.

The indicator solution consists of a 0.5 per cent aqueous solution.

Catechol Violet This indicator, also termed Pyrocatechol -Violet, is
catechol sulphonphthalein (VIII). It also possesses acid-base indicator properties
(H4D). An aqueous solution of Catechol Violet is coloured yellow; at a pH below
1-5, the colour is red; it is yellow between pH = 2 and 6 (anion HjD"), at pH = 7
It IS violet (anion HjD^ ~), and above pH = 10 the colour is blue (anion D'^ “). The
TOlour change is ascribed to the progressive ionisation of the hydroxyl groups.
The blue, strongly alkaline solutions are unstable and lose their colour fairly
rapidly, probably owing to atmospheric oxidation.

275

X, 28 QUANTITATIVE INORGANIC ANALYvSlS

OjS-0

tVlII)

Catechol violet forms colotircd compounds (usually blue or prccn-hluc) with
many metals; the most stable of these complexes arc formed in the pH rangc2-6,
so that there is a sharp colour cluinpc from yellow to blue when certain cations
(c.g., of bismuth and thorium) arc added to the indicator solution. Complexes of
the indicator with ions of divalent metals, such as Cu. Zn, Cd. Ni. and Co, do not
form until the pH is about 7, so that on adding these metal ions to the Indicator
there is only a change from violet to blue, which is less ca.sy to detect. The
determination of copper m the presence of a pyridine bufier is, however, fairly
satisfactory; the colour change is from blue to green or yclloxvisl'.-grccn.

The itulicalor solution is prepared by dissohinc 0.1 p of thedyestufTin 1 00 cm’
of water. The solution is stable for several weeks.

Bromopyrogallol Red. This metal ion indicator is dibromopyropallol
sulphonphthaleind.Xland is more resistant to oxidation than Catechol Violet; it
also po.ssc.sses acid-base indicator propertic.s. The indicator is coloured oranpe-
yellow in strongly acidic solution, claret red in nearly neutral solution, and violet
to blue in basic solution. The dye.stutT forms coloured complexes nith many
cation.s. It is valuable for the determination, inter alio, of bismuth (pH - 2-3,
nitric acid solution; end-point blue to claret red).

Hr «"on

v>-,/

\ ! \

oil

O.S-O Hr

(IX)

The intlicator solution is prepared by dissolving 0.05 g of the solid reagent in lOO
cm’ of 50 per cent ethanol,

Xylcnol Orange, This indicator, prepared by the condensation of o-
pcsolsulphonephlhalcin (Cresol Red) with formaldehyde and iminodiacetic acid,
is3,3'-/>/.s(A',A'-di(carhoxymcthyl)-aminomcthvl]-{)-cresolsulphoncplilhalcin(X).
This dycsluIT retains the acid-base propcrtics'of Cresol Red and displays metal
indicator properties even in acid solution (pH — 3-5). Acidic solutions of the
indicator arc coloured lemon-yellow and those of the metal complc.xcs intensely
red.

276

TITRIMETRIC ANALYSIS X, 28

Direct EDTA titrations of Bi, Th, Zn, Cd, Pb, Co, etc., are readily carried out
and the colour change is sharp. Iron(III) and, to a lesser extent, aluminium
interfere. By appropriate pH adjustment certain pairs of metals may be titrated
successfully in a single sample solution. Thus bismuth may be titrated at pH = 1-
2, and zinc or lead after adjustment to pH = 5 by addition of hexamine.

The indicator solution is prepared by dissolving 0.5 g of Xylenol Orange in 100
cm^ of water.

Thymolphthalein Complexone. (Thymolphthalexone). This is thymolph-
thalein di(methylimine diacetic acid) (XI);. it contains a stable lactone ring and
reacts only in an alkaline medium. The indicator may be used for the titration of
calcium; the colour change is from blue to colourless (or a slight pink).
Manganese and also nickel may be determined by adding an excess of standard
EDTA solution, and titrating the excess with standard calcium chloride solution ;
the colour change is from very pale blue to deep blue.

The indicator solution consists of a 0.5 per cent solution in ethanol.
Alternatively, a finely ground mixture (1 : 100) with A.R. potassium nitrate may be
used.

^^thylthymol Blue (Methylthymol Blue Complexone). This compound
(XII) is very similar in structure to the preceding one from which it is derived by
replacement of the lactone grouping by a sulphonic acid group. By contrast,
however, it will function in both acidic and alkaline media, ranging from pH = 6

277

TITRIMETRIC ANALYSIS X, 29

Variamine Blue. (C.I. 37255). The end-point in an EDTA titration may
sometimes be detected by changes in redox potential, and hence by the use of
appropriate redoxindicators. An excellent exampleis Variamine Blue (4-methoxy-
4'-amiiio-diphenylamine), which may be employed in the complexometric
titration of iron(III). When a mixture of iron(II) and (III) is titrated with EDTA
the former disappears first. 'As soon as an amount of the; complexing agent
equivalent to the concentration of iron(III) has been added, pFe(III) increases
abruptly and consequently there is a sudden decrease in the redox potential
(compare Section n, 24); the end-point can therefore be detected either
potentiometrically . or with a redox indicator. The stability constant of the
iron(III) complex FeY“ (EDTA = Na 2 H 2 Y) is about 10^* and that of the iroh(II)
complex FeY^" is approximate calculations show that the change of redox
potential is about 600 millivolts at pH = 2 and that this will be almost
independent of the concentration of iron(II) present. The jump in redox potential
will also be obtained if no iron(II) salt is actually added, since the extremely
minute amount of iron(II) necessary is always present in any ‘pure’ iron(III) salt.

The visual detection of the sharp change in redox potential in the titration of an
ironjlll) salt with EDTA is readily made with Variamine Blue as indicator.

The almost colourless leuco form of the base (a) passes upon oxidation into the
strongly coloured indamine (b). When titrating iron(lll) at a pH of about 3 and
the colourless hydrochloride of the leuco base is added, oxidation to the.violet-
blue compound (b) occurs with the formation of an equivalent amount of iron(II).
At the end-point of the EDTA titration, the small amount of iron(II) formed when
the indicator was introduced is also transformed into the Fe(III)-EDTA complex
FeY , whereupon the blue indamine is reduced back to the leuco base (a).

The indicator solution is a 1 per cent solution of the base in water.

A. 3 THEORY OF PRECIPITATION TITRATIONS

X, 29. PRECIPITATION REACTIONS. The most important precipitation
processes in titrimetric analysis utilise silver nitrate as the reagent (argentimetric
processes). Our discussion of the theory will therefore be confined to
argentimetric processes; the same principles can, of course, be applied to other
precipitation reactions. Let us consider the changes in ionic concentration which
occur during the titration of 100 cm^ of O.lM-sodium chloride with O.lM-silver
nitrate. The solubility product of silver chloride at the laboratory temperature is
1.2 X 10 The initial concentration of chloride ions, [Cl“], is 0.1 mole per dm^,

~ ^ Section n, 16). When 50 cm^ of O.lM-silver nitrate have been
added, 50 cm^ of O.lM-sodium chloride remain in a total volume of 150 cm^ • thus
lU ] = 50x0.1/150 = 3.33 x 10-^orpCl- = 1.48. With 90 cm^ of silver nitrate
solution [C1-] = 10 X 0.1/190= 5.3 x 10-^ or pCl" = 2.28.

279

X, 29 QUANTITATIVE INORGANIC ANALYSIS

Now xo„-^ (Ag^]y.(Cr j r. J.2>: JO'’" r.- ^

or pAp* +pCr =- 9.92 = pAgCI

In the last calculation. pCI" ■== 1.4S. hence pAp" - 9.92 -1,48^^ 8,44. In thi
manner, the vaiious concentrations of chloride and silver ions am be coinniitwt
up to the equivalence point. At the cc|uivalcncc point;

•Ag = Cl ■" ,\r<1

pAg' pCr ~ 'pAgCl 9.92/2 4.96

is pre^n'r'''*^^* soluii(m of silver chloride with no e.xcess of .silver or chloride ions

With 100,1 cm' ofsilvernitratc,solution.|Ag‘]-^ 0 I yO I - S v in-s
pAg • 4.30; pCl ■ - pAgCI - pAg ‘ - 9.92 - 4 30 =- ~ ^

TnblcX.9. Tilrationof IOOrntN)ro.lAf-.\i,aand l()0cm\.r0 lAf.

KI respcclndv with ai/.f-,\gNO. .. 1.2> I0-“'- K '

==•1.7x10 «-> lu .

Cm’ of0.Ml.

ArNO,

99..S

99.K

99,9

100.0

100.1

100,2

1(X).5

ini

102

105

110

Titration of OiIorMr

Ulralion

of Wide

per

pAp'

Pl'

pAp*

1 n

I S

K 4

I 5

2.3

7 6

2 3

J.V5

2,0

.vO

7 3

2.6

1}:

12 K

3.3

6 (i

3 3

1 2.5

3 7

6 2

3 7

12. 1

•1.0

a,3

5 6

4 0

4 3

7 9

n s
n,5
T.‘>

5,6

6 3

4 ,S

4 0

3 6

It 5

II s

122

4 0

6.(!

3 3

I 2 s

3 3

3.0

2.6

I2S

I. '.2

** #.

7.6

13 5

•v-n

nitrate are collcctcd^in^TablcV's^^^^^^^ Addition of 1 lOcm^ ofO.l A/-si!vcr

m Table X. 9. Smi.iar values for the titration of 100 cm^ of

Ihc solution; the nciual conccnlniion is .• i will contribute silver and chloride ions to

iSErc.itcrtli.in 10 times this v.ilue ic >10 ''V ' ™ R the c.sccssor.silvcr ions added

conccnirnlion produccti bv the dissnls,.s i '■ *r<a* R’e error introduced by neglecting the ionic

ensuing di.scussion. * ^ lalcn as ncghpiblc for the purpose of the

280

TITRIMETRIC ANALYSIS X, 30

0 IM-potassium iodide with O.lM-silver nitrate are included in the same table

It will be seen by inspecting the silver-ion exponents in the neighbourhood ol
the equivalence point (say, between 99.8 and 100.2 cm^) that there is a marked
change in the silver-ion concentration, although the change is more pronounced
for silver iodide than for silver chloride, since the solubility product of the latter is
about 10® larger than for the former. TTiis is shown more clearly in the titration
curve in Fig. X, 10, which represents the changes of pAg^ in the range between 10
per cent before and 10 per cent after the stoichiometric point in the titration of
O.lM-cMoride and O.lM-iodide respectively with O.lM-silver nitrate. An almost
identical curve is obtained by potentiometric titration using a silver electrode (see
Section XIV, 25); the pAg"® values may be computed from the e.m.f. figures
exactly as in the calculation of pH.

Fig. X, 10 Tifrafion curves of 100 cm^ of 0. 1 M-NaQ and of 100 cm^ of O.IM-KI
respectively with O.lM-AgNOj (calculated)

X, 30. DETERMINATION OF END-POINTS IN PRECIPITATION
reactions. Many methods are utilised in determining end-points in these
reactions, but only the most important will be mentioned here.

A. Formation of a coloured precipitate. This may be illustrated by the
Mohr procedure for the determination of chloride and bromide. In the titration
of a neutral solution of, say, chloride ions with silver nitrate solution, a small
quantity of potassium chromate solution is added to serve as indicator. At the
end-point the chromate ions combine with silver ions to form the sparingly

soluble, red silver chromate.

The theory of the process is as follows. We have here a case of fractional
precipitation (Section n, 10), the two sparingly soluble salts being silver chloride
1.2 X 10“^°) and silver chromate (K^„^ 1.7 x 10“ ^^). Let us consider an actual
^ ^“countered in practice, viz., the titration of, say, O.lM-sodium chloride
With 0. 1 M-silver nitrate in the presence of a few cm^ of dilute potassium chromate

281

X, .^0 QUANTITATIVE iNORGANIC ANALYSIS

solution. Silver chloride is the less soluble salt and, furthermore, the initial
chloride concentration is hiph. hcncx- silver chloride will be precipitated. At the
first point where red silver chromate is just precipitated, we shall have both s.ilts
in ctpiilibrium with the solution, hence;

jcr

xfCl-)

xl0-’«

'tO,'-]

“ Aj;V<0. ~

1.7x10"''

/fsV) Vc.l'rO,

(crj

V iCrO,*’-]

* *1

1.2 X 10’-'"

r V.-. 9.2 X 10" •*

V I.7y 10" '•

icri-

\ ' Aft : Atn -■ I -1

,y 10"'. Ifsilvcr chromate is to

precipitate at ihi.s chloride-ion concentration:

K:rO.=

I (Cl-'l f l >
lo- 'l '19.:^ i(r '/

I.-l X 10"*

or tlie potassium chromate solution should be 0.014, \f. It should be noted that a
slight c-vccssofsilver nitrate solution must be added before the red colour ofsilvcr
chromate is visible. In practice, a more dilute solution (0.003-0.005Af) of
potassium chromate is pencraiiv iwcd.sincx a chromate solution of concentration
0.01-0.02, \f imparts a distinct deep orange colour to the solution, which renders
the detection of the first appearance of silver chromate somewhat difficult. We
can readily calculate the error llicrcby introduced, for if |Cr(),,*~) r-- (say) 0.00.'.
silver chromate will be precipitated when:

[Ag*]

V Af <

CrCb*

/)_7 V lo' f-
V ’ 3 .V 10

2.4 X !0"'

If the theoretical concentration of indicator is used :

(Ag'i

/L7y, i0

\/ T4V1O -

1.1 X lO -

The difference is 1.3x10"' ctiui valent dm ‘ '. If the volume of the solution at the
equivalence point is 150 cm', then this corrc-sponds to 1.3 >' 10"' x. 150
X lOVlOOO = 0.02 cm-' of O.lAf-silvcr nitrate. This is the theoretical titration
error, and is therefore negligible. In actual practice another factor must be
considered, viz., the small excess of silver nitrate solution which must be added
before the eye can detect the colour chance in the solution; this i.s of the order of
one drop or oj. O.OS cm' of 0. 1 A/-st)ver nrtratc.

The titration error will increase wiih incrcasinc dilution of the .solution being
titrated and is quite appreciaiilcfn/. 0.4‘ per centfin dilute. sav.O.Ol A/, solutions
vyhen the chromate concentration is of the order 0.(X13~0,005Af. Tliis is most
simply allowed for by ipcans of an indicator blank determination, c.g., by
measuring the volume of standard silver nitrate solution required to give a

• The errors for 0 1 .Vf- .-imt 0.01 ,V/-bromidc may lx: catciil.ucd lo N 0.04 .aiwl 0 4 per cent respectively.

TITRIMETRIC ANALYSIS X, 30

perceptible coloration . when added to distilled water containing the same
quantity of indicator as is employed in the titration. This volume is subtracted
from the volume of standard solution used.

It must be mentioned that the titration should be carried out in neutral
solution or in very faintly alkaline solution, i.e., within the pH range 6.5-9. In acid
solution, the following reaction occurs:

2Cr04^- +2H+ ^2HCr04- ^Cr207^" +H 2 O

HCr 04 “ is a weak acid, consequently the chromate-ion concentration is reduced
and the solubility product of silver chromate may not be exceeded. In markedly
alkaline solutions, silver hydroxide {K^oi 2.3x10“®) might be precipitated. A
simple method of making an acid solution neutral is to add an excess of pure
calcium carbonate or sodium hydrogen carbonate. An alkaline solution may be
acidified with acetic acid and then a slight excess of calcium carbonate is added.
The solubility product of silver chromate increases with rising temperature; the
titration should therefore be performed at room temperature. By using a mixture
of potassium chromate and potassium dichromate in proportions such as to give
a neutral solution, the danger of accidentally raising the pH of an unbuffered
solution beyond the acceptable limits is minimised; the mixed indicator has a
buffering effect and adjusts the pH of the solution to 7.0 ±0.1. In the presence of
ammonium salts, the pH must not exceed 7.2 because of the effect of appreciable
concentrations of ammonia upon the solubility of silver salts. Titration of iodide
and of thiocyanate is not successful because silver iodide and silver thiocyanate
adsorb chromate ions so strongly that a false and somewhat indistinct end-point
is obtained.

B. Formation of a soluble coloured compound. This procedure is
exemplified by the method of Volhard for the titration of silver in the presence of
free nitric acid with standard potassium or ammonium thiocyanate solution. The
indicator is a solution ofiron(III) nitrate or ofiron(III) ammonium sulphate. The
addition of the thiocyanate solution produces first a precipitate of silver
thiocyanate 7.1 x 10“

Ag+ + SCN“:^AgSCN

When this reaction is complete, the slightest excess of thiocyanate produces a
reddish-brown coloration, due to the formation of a complex ion;*

Fe®+ -t-SCN“ :?i[FeSCN]2 +

'^is method may be applied to the determination of chlorides, bromides, and
iodides in acid solution. Excess of standard silver nitrate solution is added, and
the excess is back-titrated with standard thiocyanate solution. For the chloride
estimation, we have the following two equilibria during the titration of excess of
silver ions;

Ag+-fCl“^AgCl

Ag+-hSCN“:^AgSCN

This IS the complex formed when the ratio of thiocyanate ion to iron(III) ion is low; higher

complexes, [Fe(SCN)2]+,etc., are important only at higher concentrations of thiocyanate ion.

283

X, 30 QlJANTITATIVn INORGANIC ANALYSIS

The two sparingly soluble salts will be in equilibrium with the solution, hence;

1 .... rr 169

(SCN-p A''jArws 7JxlO-'-'

Wlicn llic excess of silver has reacted, the thiocyanate may rc;ict with thesilver
cliloricic. since silver tliiocyanate is the less soluble .salt until the ratio
fCrj/(SCN‘ J in the solution is 169;

AgCl + SCN " Ap.SCN 9- Cl '

This will take place Ix-fore reaction occurs with the iron(Ill) ions in the solution,
and there will consequently be a considerable titration error. It is therefore
absolutely necessary to prevent the reaction betsveen the thiocyanate and the
silver chloride. This may be effected in several ways, of « hich the first is probtibly
the most reliable:

(i) The silver chloride is filtered off before back-titrating. Sinceat this since the
precipitate will be contaminated u itli adsorbed silver ions, the suspension .sliould
be boiled for a few minutes to coagulate the silver chloride and thus remove most
of the adsorbed silver ions from its surface before filtration. The cold filtrate is
titrated,

(ii) After the addition of the silver nitrate, potassium nitrate is added .is
coagulant, the suspension is boiled for about ,3 minutes, cooled and then titrated
immediately. Desorption of .silver ions occurs and. on cooling, re-adsorption is
largely prevented by the presence of potassium nitrate.

(iii) An immi.scihlc liquid is added to ‘coat’ the silver chloride pnrticlc-s and
thereby protect tliem from interaction with the thiocyanate. The most successful
liquid is nitrobcn?cnc (en. 1.0 enr' for each 50 mg of chloride): the suspension is
well shaken to coagulate the precipitate before baek-titration.

With bromides, we have the equilibrium:

(Br ] ^ 3.5 It)

ISCN-)"

The titration error is small, and no difbcjltics arise in the determination of the
end-point. Silver iodide {K^,, 1,7 >• ' i "q is less soluble than the bromide: the
titration error is negligible, but i»ie ironflll) indicator should not be added until
c.xcess of silver is present, since the dissolved loditle reacts with Fc’' * ions:

2Fc^*+2r =i2Fc='’

C Use of adsorption indicators. K. Fajans introduced a useful type of
indicator for prempiiaiion reactions as a result of his studies on the nature of
ad.sorptioii flicaction of these indicators is due to the fact that at the equivalence
point the indicator is adsorlxrd by the precipitate, and during the process of
adsorption a change occurs in the indicator winch leads to a substance ofdiFerent
colour; they have therefore been termed ndsorptinn indicators, Tlie substances
employed arc either acid dyes, such as those of the fluore.sccin series, e.g.,
fluorescein and cosin which arc utilised as tlic sodium salts, or basic dyes, such as
tliose of the rhodaminc scric.s (e.g., rhodaminc 6G), which are applied as the
halogen salt.s.

The theory of the action of these indicators is based upon the properties of
colloids. Section XT, .1. When a chloride solution is titrated with a solution of

TITRIMETRIC ANALYSIS X, 30

Na”"

N03'

Fig. X, 11 (a) AgCl precipitated in the presence of excess of Q
(A) AgCl precipitated in the presence of excess of Ag'*'

silver nitrate, the precipitated silver chloride adsorbs chloride ions (a precipitate
has a tendency to adsorb its own ions); this may be termed the primary adsorbed
layer, and it will hold by secondary adsorption oppositely charged ions present in
solution (shown diagrammatically in Fig. X, 1 1, a). As soon as the stoichiometric
point is reached, silver ions are present in excess; these will now be primarily
adsorbed, and nitrate ions will be held by secondary adsorption (Fig. X, 11, b). If
fluorescein is also present in the solution, the negative fluorescein ion, which is
much more strongly adsorbed than the nitrate ion, is immediately adsorbed, and
will reveal its presence on the precipitate, not by its own colour, which is that of
the solution, but by the formation of a pink complex of silver and a modified
fluorescein ion on the surface with the first trace of excess of silver ions. An
alternative view is that during the adsorption of the fluorescein ion a
rearrangement of the structure of the ion occurs with the formation of a coloured
substance. It is important to notice that the colour change takes place at the
surface of the precipitate. If chloride is now added, the suspension will remain
pink until chloride ions are present in excess, the adsorbed silver will then be
converted into silver chloride, which will then primarily adsorb chloride ions. The
fluorescein ions secondarily adsorbed will pass back into solution, to which they
impart a greenish-yellow colour.

The following conditions will govern the choice of a suitable adsorption
indicator:

(a) The precipitate should separate as far as possible in the colloidal condition.
Large quantities of neutral salts, particularly of multivalent ions, should be
avoided owing to their coagulating effect. The solution should not be too
dilute, as the amount of precipitate formed will be small and the colour
change far from sharp with certain indicators.

(h) The indicator ion must be of opposite charge to the ion of the precipitating
agent. - . -

(c) The indicator ion should not be adsorbed before the particular compound
has been completely precipitated, but it should be strongly adsorbed
immediately after the equivalence point. The indicator ioii should not be too
strongly adsorbed by the precipitate; if this occurs, e.g., eosin (tetrabromo-
fluorescein) in the chloride-silver titration, the adsorption of the indicator
ion may be a primary process and will take place before the equivalence
point.

A disadvantage of adsorption indicators is that silver halides are sensitised to
the action of light by. a layer of adsorbed dyestuff. For this reason, titrations

285

X, 30 QUANTlTAniVE INORGANIC ANALYSIS

should be carried out with a minimum exposure to sunlight. When using
adsorption indicators, only 2 x 10"'' to 3 x 10"^ mol of dye per mol of silver
halide is added; thi.s small concentration is used so that an appreciable fraction of
the added indicator is actually adsorbed on the precipitate.

For the titration of chlorides, Iluoresccin may be used. This indicator h a vet)’
wcah acSd (K„ -- cn. 1 x 10' **), hence even a small amount of other acid.s reduces
the already minute ionisation, thus rendering the detection of the end-point
(which depends essentially upon the adsorption of the free anion) either
impossible or didicidt to observe. The optimum pH range is between 7 and 10.
Dichlorofluoresccin is a .stronger acid and may lx: uiili.scd in slightly acid
solutions of pH greater than 4.4 ; this indicator has t he further advantage that it is
applicable in more dilute solutions.

Eosin (tetrabromofluoresccin) is a stronger acid than dichlorofluoresccin and
can be used down to a pH of 1-2; the colour change is sharpest in an acetic add
solution (pf 1 < 3). Eosin is so .strongly adsorbed on silver halides that it cannot be
u.sed for chloride titraiions; this is because the co.sin ion cm compete with
chloride ion before the equivalence point and thereby gives a premature
indication of the end-point. With the more strongly adsorbing ions. Br", I " and
SCN the competition is not .serious and a very .sharp end-point is obtained in
the titration of these ions, even in dilute solutions. The colour on the precipitate is
magenta. Rose Bengal (dichloroteiraiodofluoresccin) and dimeihyldiiodofiuor-
cscein) have Isecn recommended for the titration of iodides.

Many other dyestuITs have been recommended as adsorption indicators, not
only for the lilratioit of halides but also for other ions. Thus cyanide ion may be
titrated witli standard silver nitrate solution using diphenylcarba7idc as
adsorption indicator (see Section X', 22|: the precipitate is pale violet at the end-
point. A selection of adsorption indicators, their properties and uses is given in
Table X. 10.

D. Turbidity method. The appearance of a turbidity is sometimes utilised
to mark the end-point of a reaction, .as in Liebig's method for cyanides (see
Section X, 22), A method whicli siioiild be incliulcd here is the turbidity procedure
for the determination of .silver uith chloride, first introdua-d by Cay Lussac. A
standard solution of sodium chloride is titrated wiiii :i solution of silver nitrate or
rice versa. Under certain conditions the addition of an indicator is unneces.sary.
because the presence of a turbidity caused by theaddition ofa few drops of one of
the solutions to the other will show that the end-point ha.s not been reached. Tlie
titration is continued until the addition of the appropriate solution produces no
turbidity. Accurate results arc obtained.

The procedure may be illustrated by the following simple experiment, which is
a modification of the Gay Lussac-Stas method. The sodium chloride solution is
added to the silver .solution in the presence of free nitric acid and a small quantity
of pure barium nitrate (the latter to assist coagulation of the precipitate). Weigh
out accurately about 0.40 g of .silver nitrate into a wcll-stoppcrod 200 cm-' bottle.
Add about I()0 cm^ of water, a few drops of concentrated nitric acid, and a small
crystal of barium nitrate. Titrate with standard 0. L\f sodium chloride by adding
20 cm at once, stoppering the bottle, and shaking it vicorously until the
precipitate of silver chloride has coagulated and settled, leaving a clear solution.
The volume of sodium chloride solution taken should leave the silver still in
excess. Continue to add the chloride solution, I an ’ at a time, stoppering and
shaking after each addition, until no turbidity is produced: note the total volume

286

Table X, 10. Selected adsorption indicators: properties and uses

TITRIMETRIC ANALYSIS X, 30

rj;

n cq

c C3

.5 1 >-■

4_. ’-1 3

M M O
®

W o.<

t T •?

^ *5

C/D

I 60

^ S

o 2
o

td g

■3 Ci-

a ^

5:1 «
^ •=
c/3

C ^

- o <a
a-.= ±3

.52 o t

2 'a I

f 7" 7 « I

I I a s .1 ^

C Q d O

CCS

6 s

s ^

5 +

S 1 XI

I " 0.

0 :S

+ s 'S

^ .2^

1 cd ”♦

z.^o

U V 5

60'-' + r-i

< J= 60 w

£ '5 ^

•5 ^ S3 U

? + <0 'rf

<, 1-^

u O
O
3

o
3
C

Id ^
w 60 '-r

o « &»
cd ® ■—
i- 4) 2

S o.§

a Pi o
O 'S '7',S
*2 W O N

JS ^ O ^

O O '7 v-«

S la S H

* 3

a8
o- c

c 2

^ o

w (l> *0
Z e0r2

><-s>s 3 C

^ O S
*^ *-• 52

m p o

00 s.

O o CO
*- O

O. a> t

3 t*% «n

s-iig

^ >!5 w

^ “ >>si
C Q-’c

•:: e- o <N

CO cd ■«-> O
ii o

CQ c/3

‘>

•a

C I

? ;>.

g;a

3'0
.3 u

T 1

2 A. Q, Ci.

w) a G,

a ^
. T3 §

*" O M

O « m

I JS i

.ts *-
03 ^ CQ

•B < B

’B m '$

+ , •*+
601 m 60

<U<

C/3 S
O

N *3
33 V

cj »- .-5

c: G B
§ *C «

^ cd *0

^ .a ®

o — ^ *c J3
03 ^Di Oi

287

X, 31 QUANTITATIVE INORGANIC ANALYSIS

of sodium chloride solution. Repeat the determination, using a fresh sample of
.silver nitrate of about the same weight, and run in initially that voumcofthcO.lM
sodium chloride, less 1 cm-*, which ihefir.si titration has indicated will be required,
and thereafter add the chloride solution dropwise (i.c.. in about O.O.S cm^
portions). It will be found that the end-point can be determined \vithin one drop.

A detailed account of modern nephelometric tcchnicjues is given in Chapter
XX (Ncphelomctry and Turbidimetry).

A. 4 THEORY OF OXIOATION-REOUCTION TITRATIONS

X, 31. CUANGK OF THi: KLF.CI'RODi: FOTF.NTIAI. DURING THE
rn’RAriON of a KFDUCI ant with an oxidant, in Sections X.

11-16 it has been shown how to calculate the change in pH during acid--b.asc
titrations, and how the titration curves thus obtained can be u.scd (rr) to ascertain
the most suitable indicator to he used in a given titration, and (h) to compute the
titration error. Simdar proccdurc.s may be carried out for oxidation- reduction
titrations, and we will consuler fir.si a simple ease which involves only the valency
change of ions, and is theoreiicaliy independent of the hydrogen-ion
concentration. A suitable example, for purposes of illustration, is the titration of
100 cm-’ off), liV-iron(ll) with 0. 1 A'-cerinm(lV) in the presence of dilute sulphuric
acid:

Ce"*-fFc” =:tCc''' -f Fe’*

The quantity corresponding to |H’) in acid -base titralion.s is the ratio 10.x],/
[Red]. We arc concerned here with two systems, the Fe-' '/Fc* ' ion cla'trodc(l).
and the Cc‘‘ ‘ /Cc-' * ion electrode (2).

For(l)at 25 C;

,.o 0.0591, {Fc-''j IFc’-*!

F, =:FF+ -.--log-- .«-0,75 + 0.0.‘;911ogL--.~.

■(Fc=') '

For (2). at 25 C:

P ,-^*^0.0591 lev-)

Fj = /: Y -f — log ~ 4

RCe

4- 1.45 4- 0.0591 log

..W,

Tlie equilibrium constjint of the reaction is given by (.Section H, 25):
log A =

~0^5^
= 11.84

(1.45-0.75)

or K = 7x 10"

The reaction is therefore virtually complete.

Dunng the addition of the cerium(l V) solution up to the equivalence point, its
only cITcct will be to o.xidi.sc the iron(n)(since K is large) and consequently change

288

TITRIMETRIC ANALYSIS X, 31

the ratio [Fe^ ■*']/[Fe^ When 10 cm^ of the oxidising agent have been added,
[Fe^'^]/[Fe^'^] = 10/90 (approx.), and = 0.75 +0.0591 log 10/90 =

0.75-0.056 = 0.69 volt;

with 50 cm^ of the oxidising agent, = Ef = 0.75 volt
with 90 cm^ El = 0.75 + 0.0591 log 90/10 = 0.81 volt
with 99 cm^ El = 0.75 = 0.059 1 log 99/1 = 0.87 volt
with 99.9 cm^ Ei = 0.75+0.0591 log99.9/0.1 = 0.93 volt

At the equivalence point (100.0 cni^)[Fe^.'^] = [Ce^’'']and[Ce‘*^''']= [Fe^"^], and
the electrode potential is given by;*

Ef+Ef 0.75 + 1.45
2 ” 2

1.10 volts

The subsequent addition of cerium(IV) solution will merely increase the ratio
[Ce'‘*]/[Ce3+].Thus;

with 100.1 cm^ Ej = 1.45 + 0.0591 logO.1/100 = 1.27 volts
with 101 cm^, El = 1.45+0.0591 log 1/100 = 1.33 volts
with llOcm^, El = 1.45 + 0.0591 log 10/100 = 1.39 volts
with 190 cm^, Ej = 1.45 + 0.0591 log90/100 = 1.45 volts

These results are shown in Fig. X, 12.

It is of interest to calculate the iron(II) concentration in the neighbourhood of
the equivalence point. When 99.9 cm^ of the cerium(IV) solution have been
added, [Fe^*] = 0.1 x 0.1/199.9 = 5 x 10"^ or pFe^* = 4.3. The concentration

Fig. X, 12 Titration of 100 cm^ of 0.1 M-lron(n) >vith 0.1 Af-cerium sulphate (calculated)

* For a deduction of this expression and a discussion of the approximations involved, see a textbook
of electrochemistry. It can similarly be shown that for the reaction:

0 Oxi +ft Redii Oxii +0 Red|

the potential at the equivalence point is given by:

hEf + aEf
r-o

ti + b

where £f refers to Oxi, Red], and Ef to Oxii, Rcdii.

289

X, 31 QUANTITATIVE INORGANIC ANALYSIS

III the equivalence point is given by(Scciion H, 25):

(Fc^^I/(Fe’l - Jk - ^/7x T0^? -g.4 x 11)^

Now (Fe^") ■-=- 0.05;V. lienee {Fc='] 5 >: 10* ^K.Sx 10* = 6x |0~*iV, or

pFc'* "7.2. Upon the addition of 100,1 cm"' of ceriumfiV) solution, the
reduction potential {l itlcsuprii) is 1.2? volts. The {Fe’ ' ] is practically unchanecd
at 5x !0'*iV, rmd we may calculate (Fc*'j with sullicient accuracy for our
purpose from the equations;

[Fc^U

F.-.F;-a(l059llogj"p;|

Sx 10-

1.27 0,75 -‘-0,0591 log -

(Ic- i

{Fe-‘] - I X lir"-
or pFc*’ 10

TluispFe-* changes .honi 4, t to lObctwcenO.l per cent Iscforc and 0.! percent
after the stoichiometric eiui-poini. These quantities are of importance in
connection witli the use of indicators for the detection of the equivalence point.

It IS evident that the abrupt cltanec of the potential in liic neighbourhood of the
equivalence point is dependent upon the standard potentials of the two

o,vidation-rcduction systems that
are involved, and therefore upon thc
equilibrium constant of the reac-
tion; it is independent of the concen-
trations unless these are e.xtrcnicly
small. Tlic change in redox potential
for a number of typical
oxidation-reduction systems is e.x-
hibited graphically in Rg. X. 1 3, For
the Mn 04 ', Mn’^ system and
others which arc dependent upon the
pH of the solution, the hydrogen-ion
concentration is assumed to be
molar: lower acidities give lower
potentials. The value at 50 per cent
oxidised form will, of course, cor-
respond to the standard redox
potential. As an indication of the
application of the cun’cs, we may
take the titration of iron(H) with
potassium dichromatc. The titration
curve would follow that of the
Fcni)/Fe(ni) system until the end-
point was reached, then it would rise
steeply and continue along the curve
Pet 11(11 (iMiii-.c.l iWtn rorthcCr207^'‘ /Cr^ '^s.vslcni.'lhcpo-
l ig. X. !,I N'ariiition of Redox tential at the equivalence point can

I otciitials with Oxidnnt/Uciluclant Ratio be computed a,s already described.

290

TITRIMETRIC ANALYSIS X, 32

It is possible to titrate two substances by the same titrant provided that the
standard potentials of the substances being titrated, and their oxidation or
reduction products, differ by about 0.2 V. Stepwise titration curves are obtained
in the titration of mixtures or of substances having several oxidation states. Thus
the titration of a solution containing Cr(VI), Fe(III) and V(V) by an acid
titanium(III) chloride solution is an example of such a mixture; in the first step
Cr{VI) is reduced to Cr{III) and V(V) to V(IV); in the second step Fe(III) is
reduced to Fe(n); in the third step. V(iy) is reduced to V(III); chromium is
evaluated by difference of the volumes of titrant used in the first and third steps.
Another example is the titration of a mixture of Fe(II) and V(IV) sulphates with
Ce(IV) sulphate in dilute sulphuric acid: in the first step Fe(II) is oxidised to
Fe(III) and in the second ‘jump’ V(IV) is oxidised to V(V), the latter change is
accelerated by heating the solution after oxidation of the Fe(II) ion is complete.
The titration of a substance having several oxidation states is exemplified by the
stepwise reduction by acid chromium(II) chloride of Cu(II) ion to the monovalent
state and then to the metal.

X, 32. FORMAL POTENTIALS. Standard potentials are evaluated with
full regard to activity effects and with all ions present in simple form: they are
really limiting or ideal values and are rarely observed in a poteritiometric
measurement. In practice, the solutions may be quite concentrated and
frequently contain other electrolytes; under these conditions the activities of the
pertinent species are much smaller than the concentrations, and consequently the
use of the latter may lead to unreliable conclusions. Also, the actual active species
present (see example below) may differ from those to which the ideal standard
potentials apply. For these reasons ‘formal potentials’ have been proposed to
supplement standard potentials. The formal potential is the potential observed
experimentally in a solution containing equal numbers of moles of the oxidised
and reduced substances together with other specified substances at specified
concentrations. It is found that formal potentials vary appreciably, for example,
with the nature and concentration of the acid that is present. The formal potential
incorporates in one figure the effects resulting from variation of activity
coeflacients with ionic strength, acid-base dissociation, complexation, liquid-
junction potentials, etc., and thus have a real practical value. Formal potentials
do not have the theoretical significance of standard potentials, but they are
observed values in actual potentiometric measurements. In dilute solutions they
usually obey the Nernst equation fairly closely in the form:

n ^[Red]

at 25 °C

where E®' is the formal potential and corresponds to the value of E at iinit
concentrations of oxidant and reductant, and the quantities, in square
brackets refer to molecular concentrations. It is useful to determine and to
tabulate E®' with equivalent amounts of various oxidants and their conjugate
reductants at various concentrations of different acids. If one is dealing with
solutions whose composition is identical with or similar to that to ' which the
formal potential pertains, more trustworthy conclusions can be derived from
formal potentials than from standard potentials.

To illustrate how the use of standard potentials may occasionally lead to

291

X, 33 QUANTITATIVE INORGANIC ANALYSIS

erroneous conclusions, let us consider the hcxacyunorcrratcfn)-hexacyano-
rcrratc(IIl) ;md the iodide-iodine systems. The standard potcntial.s arc:

|re(CN),]^"+c=i(Fc(CN).r':

I,+2c:±2r

i:^' = -I- 0.36 volt
= +0.54 volt

It would be c.xpee!ed that iodine would quantitatively oxidise hexa-
cyanofcrratcdl) ions:

2IFe(CN),r" +1, - 2(r-c{CNW^' +21

In actual fact [Fc(CNF,l'‘" ion o.vidFcs iodide ion quantitatively in media
containing about 1 ^/-hydrochloric, .sulpliuric. or perchloric acid, Thi.s is because
in .solutions of low pil, protonation occurs and the s|>ecic,s derived from
H^FciCN),, arc weaker than iho.se derixed from 1 1 ,FcfCN)r,; the activity of the
(FefCNy " ion is decreased to :r greater extent than that of the {Fc(CN)<;f ” ion.
and therefore the reduction potential is increased. The aciimi redox potential ofa
solution containing equal concentrations of both cyanoferratc.s in Idf-HCI,
H 2 SO 4 or HCiOi is +0.71 volt, a value that is greater than the potential of the
iodine- iodide couple.

Some results of formal potential measurements may now be mentioned. If there
is no great difference in complcxation of either the oxidant or its conjugate
rcductant in various acids, the formal potentials lie close together in these acids.
Tims for the Fcfll) Feflll) .system E" ~ +0.77 volt. /:•" +0.73 volt in lAf-

HC! 04 . + 0.70 volt in I .ff-riCl.+O.dS volt in l.f/.H-SO... and + 0.61 volt in
O. 5 A/-H 3 PO 4 + l.Vf-lFSOj. It would seem that complcxation is least in
perchloric acid and crcatc.st in pho.sphoricacid.

For the CdllD-CcflV) .system E'" - +1.44 volts in l.\/-ll,S 04 . + 1.61 volts
in lAZ-HNOj, and+1.70 voll.s in lAf-HCl 04 . Perchloric acid solutions of
ccrium(IV) perchlorate, although unstable on standing, react rapidly and
quantitatively with many inorganic compounds and have greater oxidising
power than ccrium(IV) sulphate -sulphuric acid or ccriumflY) nitrate -nitric acid
solution.s.

X,33. DETECTION OF THE END-POINT IN OXIDA'nON-
REDUCTION TITRATIONS. A. Internal oxidation-reduction
indicators. Wc Iiavc alrcad) seen (Sections X, 10-16) that acid -ba.se indicators
arc employed to mark the sudden > ''angc in pH during acid-base titrations.
Similarly an oxidation-reduction lOicalor sliould mark the sudden change in
the oxid.ation potential in the neighbourhood of the equivalence point in an
oxidation -reduction titration. The ideal oxidation-reduction indicator will be
one with an oxidation potential intermediate between that of the solution titrated
and that of the titrant, and wliich exhibits a sharp, readily detectable colour
change.

An oxidation-reduction indicator (redox indicator) is a compound which
exhibits dilTcrcnt colour.s in the oxidised and reduced forms:

lno, + nc;?eIn,;,j

Theoxid.ation and reduction should be reversible. At a potential £ the ratio of the
concentrations of the two forms is given by the Nernst equation:

292

TITRIMETRIC ANALYSIS X, 33

Pnpx]

[InRcd]

where is the standard (strictly the formal) potential of the indicator. If the
colour intensities of the two forms are comparable a practical estimate of the
colour-change interval corresponds to the change in the ratio [Inoi]/[InRed] from
10 to this leads to an interval of potential of ; ■ - ■

£ ^ £0+2:2^ volts at 25 °C
n ^ .

If the colour intensities of the two forms differ considerably the intermediate
colour is attained at a potential somewhat removed from but the error is
unlikely to exceed 0.06 volt. For a sharp colour change at the end-point, £®
should differ by about at least 0.15 volt from the standard (formal) potentials of
the other systems involved in the reaction.

One of the best oxidation-reduction indicators is the 1,10-phenanthroline-
iron(II) complex. The base 1,10-phenanthroline combines readily in solution with
iron(II) salts in the molecular ratio 3 base; 1 iron(II) ion forming the intensely red
l,10-phenanthroline-iron(II) complex ion; with strong oxidising agents the
iron(III) complex ion is formed, which has a pale blue colour. The colour change
is a very striking one ;

[Fe(Ci2H8NJ3]^ + + e ^[Fe(C,2H8N2)3]^ +

pale blue deep red

The Standard redox potential is 1.14 volts; the formal potential is 1.06 volts in
IM-hydrochloric acid solution. The colour change, however, occurs at about 1.12
volts, because the colour of the reduced form (deep red) is so much more intense
than that of the oxidised form (pale blue). The indicator is of great value in the
titration of iron(II) salts and other substances with cerium(IV) sulphate solutions.
It is prepared by dissolving 1,10-phenanthroline hydrate (molecular weight
= 19ll) in the calculated quantity of 0.02M acid-free iron(II) sulphate, and is
therefore l,10-phenanthroline-iron(II) complex sulphate (known asferroin). One
drop is usually sufficient in a titration; this is equivalent to less than 0.01 cm^ of
0.1 iV oxidising agent, and hence the indicator blank is negligible at this or higher
concentrations.

It has been shown (Section X, 31) that the potential at the equivalence point is
the mean of the two standard redox potentials. In Fig. X, 12, the curve shows the
variation of the potential during the titration of O.liV-iron(II) ion with O.IW-
cerium(IV) solution, and the equivalence point is. at 1.10 volts. Ferroin changes
from deep red to pale blue at a redox potential of 1.12 volts; the indicator will
therefore be present in the red form. After the addition of, say, a 0.1 per cent excess
of cerium(IV) sulphate solution the potential rises to 1.27 volts, and the indicator
is oxidised to the pale blue form. It is evident that the titration error is negligibly
small.

The standard or formal potential of ferroin can be modified considerably by
the introduction of various substituents in the 1,10-phenanthroline nucleus. The
most important substituted ferroin is 5-nitro-l,10-phenanthroline iron(II)
sulphate (nitroferroin) and 4,7-dimethyl-l,10-phenanthroline iron(II) sulphate
(dimethylferroin). The former (£® = 1.25 volts) is especially suitable for titrations

293

X, 33 QUANTITATIVE INORGANIC ANALYSIS

usinc Cc(lV)in nitric or perchloric acid solution where the forma! potential of the
oxidant is high. The 4.7-dimclhylfcrroin has a sufficiently low formal potential
(E^ 0,88 volt) to render it u.sc'fui for the titration of Pcfll) with dichromatc in

0.5Af-sulphuricacid.

Mention should he made of one of the earliest internal indictitors. This is a 1 per
cent solution of diphenylarninc in concentrated sulphuric acid, and was
introduced by Knop for the titration of ironfll) with potassium dichromatc
solution. An intense blue-violet coloration is produced at the end-point. The
addition of phosphoric acid is desirable, for it lowers the formal potential of the
Fcfllll-Fcfll) sy.stem so that the equivalence point potential coincides more
nearly with that of the indicator. The action of tliphcnylaminc (I) as an indicator
depends upon its oxidation first into colour)c.<s diphenylbenzidinc {11), wliicli is
the real indicator and is reversibly further oxidised to diphcnyllrcnzidinc \iolct
(HI).

'—Ntl— t '■

\ /

■ (Il.t

■'(tltl

Diphcnylben7,idinc siolct undergoes further oxidation if it i-s allowed to stand
with excess of dichromatc solution; this further oxidation is irreversible, and red
or yellow products of unknown composition arc produced.

A solution of diplicny!bcn7idinc in concentrated sulphuric add nets similarly
to diphenylarninc. The reduction potential of the sy.stem 11, HI is 0.76 volt in 0.5-

1 alilc X, 1 1. Some oxid.niinn ndticlioii indicator.

tmticatot

Colour chjncr

Oiido.cd

fonn

Rcdiuxd

form

Ponnal
. jwlcntiit
Oelts)
at pll « 0

S-.\'itro-l,IO-p!icnnntIirolitic iror.iIDsutpli.iic
(nitroferroini

I’.de blue

Red

1.25

I.tO-PhetulirolincironllDsutplmicirerroin)

blue

Ret!

!,W

2,2 -Ilipynd)! ironfll) sutphruc

baim blue

Red

1,02

5,6-Djmc{h) Ifcrroin

I'a’c btvie

Red

(1,97

A -phenj trinUirrinnie add.

Purple fctl

Coloiirie-.t

0.S9

4,7-Dimc:hjl-l, tO-pticn.inihro!i,iejroi5(III

sulphate (4, T-dimeiKylfcooiiil

P.UebUie

Red

P.SS

Diplicinlammesulphoi’ic acid

Cotoutless

OSS

Diphcpylt'cnridinc

Vinlfl

Colourlc-^*!

0.76

Diphcnyl.imiiic

X'lotct

Colourless

0.76

.’.3'-Oinic(h>I»aphthidine

Purpitsh-red

Co!our!e^^

0,71

Starch-Ij , KI

Blue

Cotourtcvs

0.5?

Xfclhylcnc blue

niuc

Colourless

0,52

294

TITRIMETRIC ANALYSIS X, 33

IM-sulphuric acid. It is therefore evident that a lowering of the potential of the
Fe(III)-(Fe(II) system is desirable, as already mentioned, in order to obtain a
sharp colour change. The disadvantage of diphenylamine and of diphenyl-
benzidine is their slight solubility in water. This has been overcome by the use of
the soluble barium or sodium diphenylaminesulphonate, which are employed in
0.2 per cent aqueous solution. The redoX potential is slightly higher (0.85 volt
in 0.5M-sulphuric acid), and the oxidised form has a reddish-violet colour
resembling that of potassium permanganate, but the colour slowly disappears on
standing; the presence of phosphoric acid is desirable in order to lower the redox
potential of the system.

A list of selected redox indicators, together with their colour changes and
reduction potentials in an acidic medium, are given in Table X, 1 1.

At this stage reference may be made to potential mediators, i.e., substances
which undergo reversible oxidation-reduction and reach equilibrium rapidly. If
we have a mixture of two ions, say M^'*' and M'^, which reaches equilibrium
slowly with an inert electrode, and a very small quantity of a cerium(IV) salt is
added, then the reaction;

M++Ce‘‘+ -^M^+-bCe2 +

takes place until the tendency of M'*' to be oxidised to is exactly balanced by
the tendency of Ce^"*^ to be oxidised to Ce'^'*', that is, until the M"'" and
Ce^'*' potentials are equal. A platinum or other inert electrode rapidly
attains equilibrium with the Ce(III) and Ce(IV) ions, and will soon register a
stable potential which is also that due to the -t'e:?^M‘*' system. If the
potential mediator is employed in small amount, then a negligible quantity of M
is converted into when equilibrium is reached, and the measured potential
may be regarded as that of the original system. Potential mediators are, of course,
useful in the measurement of the oxidation-reduction potentials of redox
systems; in this connection mention may be made of the use of potassium iodide
(s iodide-iodine system) in the arsenate-arsenite system in acid solution. It is
evident that redox indicators (e.g., 1,10-phenanthroline iron(II) ion) may act as
potential mediators.

B. The reagent may serve as its own indicator. This is well illustrated by

potassium permanganate; here, however, sensitive internal indicators (1,10-
permanganate will impart a visible pink coloration to several hundred cm^ of
solution, even in the presence of slightly coloured ions, such as iron(III). The
colour of cerium(IV) sulphate and of iodine solutions have also been employed in
the detection of end-points, but the colour change is not so marked as for
potassium permanganate; here, however, sensitive internal indicators (ortho-
phenanthroline iron(II) ion or N-phenylanthranilic acid and starch respectively)
are available. ...

This method has the drawback that an excess of oxidising agent is always
present at the end-point. For work of the highest accuracy, the indicator blank
may be determined and allowed for, or the error may be considerably reduced by
performing the standardisation and determination under similar experimental
conditions.

C. External indicators. The best-known example of an external indicator
in a redox process is the spot-test method for the titration of iron(II) with
standard potassium dichromate solution. Near the equivalence point, drops of
the solution are removed and brought into contact with dilute, freshly prepared

295

X, 34 QUANTITATIVE INORGANIC ANALYSIS

potassium hcxacyanorenalcdll) solution on a spot plate. Tlic end-point is
reached when the drop first fails to give a blue coloration. Another c.xamplc is
provided by the titration of zinc ion.s with standard potassium he.xacyano-
fcrrale(II) solution; here a solution of tiranyl acetate or nitrate i.s the c.xicrnai
indicator, and titration is continued until a drop of the solution just imparts a
brown colour to the indicator. External indicators arc virtually su{>crscded by the
more satisfactory internal oxidation- reduction indieator.s: thus in the first
example l.lO-phcnaiuhroIine ironfll) ton or A'-plicnylanthranilic acid is
.suitable, whilst for the second 3. .V-dimetby!naphthidinc may be used.

1). Potcnfinmciric methods. This is a procedure which depends upon
measurement of the c.m.f. between a reference electrode and an indicator (redox)
electrode at .suitable intervals during the titration, i.c,. a potcntiomctric titration is
carried out. The procedure is discussed fully in Chapter XIV and suflice at this
stage to point out that the procedure is applicable not only to those cases where
suitable indicators arc available, but al^o to those cases, c.g.. coloured or very
dilute solutions, where the indicator method is inapplic.able, or of limited
accuracy.

B EXPERIIVIEWTAL DETAILS

B.1 AQUEOUS ACID-BASE TITRATIONS

Acidinictry and .Mkalimctry

X, 34. I’RE FARATION’ 01 A STANDARD ACil D. Disais^km. Two acids,
namely hydrochloric acid and sulphuric acid, arc wirlely cnipKned in the
preparation of .standard soitiitons of acids. Doth of these arc commercially
available as concentrated solutions; conccmratcrl hydrochloric acid is about
10.5-1 2, \f, and concentrated sulphuric acid is ;iboui \i>M. By suitable dilution,
solutions of anj de.sircd (ipproxinsoii- sticnptli may be readily prepared.
Hydrocliloric acid is generally preferred, since most chlorides arc soluble in
water. Sulphuric acid forms insoluble salt.s with calcium and Ixiriuin hydroxides;
for titration of hot liquids or for determinations which rctpiirc boiling for some
tirne with excess of acid, standard .s;ilphuric acid is. however, preferable. Nitric
acid is rarely employed, because it almost invariably contains a little nitrous acid,
which has a desiructivc action upon many indicators.

For the present, we sliall coniine our attention to the preparation of standard
.solutions of hydrocliloric acid. Two methods are available. The first utilises Ihe
experimental fact that aqueous solmion.s of hydrochloric acid lose either
hydrogen chloride or water upon boiling, acconiing as to whether they arc
stronger or weaker than the con.siant-boihng-point mixture, until they attain a
practically constant composition Iconstant-boihng-point mixture), which
depends upon the prc\ ;iiling pressure. The compositimi of this constaitt-boiling
nu.xturc and its dcpcmlcncc upon pressure iiave been determined with great
accuracy by I'oulk and I lolliiigsu ortit. The relevant data arc collected in Table X.

296

TITRIMETRIC ANALYSIS X, 35/36

Table X, 12 . Composition of constant-boiling-point hydrochloric acid

Pressure (mm of Hg)

Per cent HQ in acid
(vac. wt)

Grams of acid, weighed in ■
air, containing 36.47 g of

780

20.173

180,621

770

180.407

760

20,221

180.193

750

20.245

179.979

740

20.269

179.766

730

20.293

179.555

The constant-boiling-point acid is neither hygroscopic nor appreciably
volatile, and its concentration remains unchanged if kept in a well-stoppered
vessel out of direct sunlight. This acid may be employed directly in the
preparation of a solution of hydrochloric acid of known concentration. '

In the second method a solution of the approximate strength required is
prepared, and this is standardised against some standard alkaline substance, such
as sodium tetraborate or anhydrous sodium carbonate; standard potassium
iodate or pure silver may also be used (see Sections X, 84, 130). If a solution of an
exact normality is required, a solution of an approximate strength somewhat
greater than that desired is first prepared; this is suitably diluted with water after
standardisation (for a typical calculation, see Section X, 4).

The student should read the theoretical sections X, 1-18, before embarking
upon the experimental work.

X, 35. PREPARATION OF CONSTANT-BOILING-POINT HYDRO-
CHLORIC ACID. Mix 400 cm^ of pure concentrated hydrochloric acid with
250-400 cm^ of distilled water so that the specific gravity of the resultant acid is
1.10 (test with a hydrometer). Insert a thermometer in the neck of a 1-litre Pyrex
distillation flask so that the bulb is just opposite the side tube, and attach a
condenser to the side tube; use an all-glass apparatus. Place 500 cm^ of the
diluted acid in the flask, distil the liquid at a rate of about 3-4 cm^ per minute and
collect the distillate in a small Pyrex flask. From time to time pour the distillate
into a 500-cm^ measuring-cylinder. When 375 cm^ has been collected in the
measuring-cylinder, collect a further 50 cm^ in the small Pyrex flask; watch the
thermometer to see that the temperature remains constant. Remove the receiver
and stopper it; this contains the pure constant-boiling-point acid. Note the
barometric pressure to the nearest mm at intervals during the distillation and
take the mean value. Interpolate the concentration of the acid from Table X, 12,

X, 36. DIRECT PREPARATION OF O.IM-HYDROCHLORIC ACID
FROM THE CONSTANT-BOILING-POINT ACID. Clean and dry a small,
stoppered conical flask; a glass-stoppered flask is preferable. After weighing, do
not handle the flask directly with the fingers; handle it with the aid of a tissue or a
linen cloth. Add the calculated quantity of constant-boiling-point acid required
for the preparation of 1 dm^ of O.lM-acid (see Table X, 12) with the aid of a
pipette; make the final adjustment with a dropper pipette. Reweigh the flask to
0.001 g after replacing the stopper. Add an equal volume of water to prevent loss

297

X, 34 QUANTITATIVE INORGANIC ANALYSIS

nota^sium hcxacvanoferralcClII) solution on a spot plate. Tlic end-point is
reached when the" drop first fail.s to give a blue coloration. Another example is
provided by the titration of zinc ions with standard potassium hc.xacyano-
fcrratc{Il) sohition; here a solution of uranyl acetate or nitrate is the external
indicator, and titration is continued until a drop of the solution just imparts a
brosTO colour to the indicator. Externa! indicators arc virtually superseded by the
more sati.sfactory internal o.xidation- reduction indicators: thus in the first
example l.lO-phcnanthroline irondD ion or A'-phcnyianthranilic acid is
suitable, whilst for the second 3, 3'-dimcthylnap!ili)idinc may be used.

D, Potcntiomelric mcthmls. This is a procedure which depends upon
measurement of the c-m.f. between a reference electrode and an indicator (redox)
electrode at suitable intervals during the titration, i.c., a potcntiomctric titration is
carried out. The procedure is discussed fully in Chapter XIV and suffice at this
stage to point out that the procedure is applicable not only to tho.se casc,s where
suirabic indicators arc avail.ible. but rd.sr? to Iho.se cases, c.g., coloured or scry
dilute solutions, where the indicator method is inapplicable, or of limited
accuracy.

B EXPERIMENTAL DETAILS

B.1 AQUEOUS ACID-BASE TITRATIONS

Acidimetry and .Mkalinietry

X. 34. PRI-I'ARATION OF A STANDARD ACID. Dbeinsimt. Two acids,
namely hydrochloric acid and sulphuric acid, arc widely employed in the
preparation of .standard solutions of acids. Roth of these arc commercially
available as concentrated soluiions; eoiicentraied hydrochloric acid is about
10.5 -12.'/, and concentrated .sulphuric acid is about ISA/. By .suitable dilution,
solutions of any desired up/’rov/wafe sifength may lx; readily prt'pared.
Hydrochloric acid is generally preferred, since most clilorides arc soluble in
water. Sulphuric acid forms insoluble salts with calcium and barium hydroxides:
for titration of hot liquids or for determinations vvliieh require boiling for some
tunc with excess of acid, standard .suiphuric acid is, however, preferable. Nitric
acid is rarely employed, bcc.iuseit almost inxariably contains a little nitrous acid,
wiiich has a destructive action upon many indicators.

For tlic prcsciu. wc shall confine our attention to the preparation of standard
solutions of hydrochloric acid. Two methods are available. The first utilises the
experimental fact tliat aqueous solutions v'f hvdrochloric acid lose either
hydrogen chloride or water upon boiling, ticcording as to whether they are
.stronger or weaker th:tn tlic consiani-hoiling-poiiU mixture, until they att.iin a
practically constant conipoMiion (constanl-boilinc-poinl mixture), which
depends upon the prevailing prcs.surc. The composition of lliis constant-boiling
mixture and its dependence upon prc.ssurc have been deiermineti with grc:it
accuracy by Foulk and Hollingsworth. The relevant data arecollected in Table X.

296

TITRIMETRIC ANALYSIS X, 35/36

Table X, 12. Composition of constant-boiling-point hydrochloric acid

Pressure (mm of Hg)

Per cent HCl in acid
(vaa wt)

Grams of acid, weighed in
air, containing 36.47 g of

HCl

780

20.173

180.621

770

20.197

180.407

760

20,221

180.193

750

20.245

179.979

740

20.269

179.766

730

20.293

179.555

The student should read the theoretical sections X, 1-18, before embarking
upon the experimental work.

X, 35. PREPARATION OF CONSTANT-BOILING-POINT HYDRO-
CHLORIC ACID. Mix 400 cm^ of pure concentrated hydrochloric acid with
250-400 cm^ of distilled water so that the specific gravity of the resultant acid is
1.10 (test with a hydrometer). Insert a thermometer in the neck of a 1-litre Pyrex
distillation flask so that the bulb is just opposite the side tube, and attach a
condenser to the side tube; use an all-glass apparatus. Place 500 cm^ of the
diluted acid in the flask, distil the liquid at a rate of about 3-4 cm^ per minute and
collect the distillate in a small Pyrex flask. From time to time pour the distillate
into a 500-cm^ measuring-cylinder. When 375 cm^ has been collected in the
measuring-cylinder, collect a further 50 cm^ in the small Pyrex flask; vvatch the
thermometer to see that the temperature remains constant. Remove the receiver
and stopper it; this contains the pure constant-boiling-point acid. Note the
barometric pressure to the nearest mm at intervals during the distillation and
take the mean value. Interpolate the concentration of the acid from Table X, 12.

X, 36. DIRECT PREPARATION OF O.IM-HYDROCHLORIC ACID
FROM THE CONSTANT-BOILING-POINT ACID. Clean and dry a small,
stoppered conical flask; a glass-stoppered flask is preferable. After weighing, do
not handle the flask directly with the fingers ; handle it with the aid of a tissue or a
linen cloth. Add the calculated quantity of constant-boiling-point acid required
for the preparation of 1 dm^ of O.lM-acid (see Table X, 12) with the aid of a
pipette; make the final adjustment with a dropper pipette. Reweigh the flask to
0.001 g after replacing the stopper. Add an equal volume of water to prevent loss

297

X. 37 QUANTITATIVE INORGANIC ANALYSIS

of acid, and iransferthc contents to a I dm^ graduated flask. Wash out the weigh-
ing fiask several times with distilled water and add the washings to the original
solution. Make up to the mark with distilled water. Insert the stopper and mix the
solution thoroughly by shaking a.nd inverting the flask repeatedly.

Note. Unicss a' solution of exact concentration is required, it is not
necessary to weigh out tlic exact quantity of constant-boiling acid; the
concentration may be calculated from the weight of acid used. Thus, if 1 8.305 g of
acid, prepared at'760 mm, was diluted to 1 dm\ its concentration would be
18.305/180.193 = 0.1015S.\/ dm~\

X,37. PRi:PARATIO.NOFAPRROXl.MATKLY0.1Af-HYDROCIlLORIC
ACID AND STAN'DARDLS.ATION. .Measure out by means of a graduated
cylinder or a burette 9 cm' of pure concentrated hydrochloric acid: pour the acid
into a litre rnca.suring-cylinder containing about 5(X)cm' of distilled water. Make
up to the litre mark with disiilled water and thoroughly mix by .shaking. This will
give a solution approximately 0. i Sf 1 1 1

.Note. 1. If 1 .M-h\drochloric acid is required, use 90 cm'* of the
concentrated .acid. If 001 Al-.acid is re-quired. dilute two 50-cm’ portions of the
approximately 0. 1 .M-.add. remoxed with a 50-cm-' pipette, in a graduated flask to
1 litre.

Approximately (0,05. Mi-s»i!phuric acid is .similarly prepared from 3 cm^ of pure
concentrated .vulplmric acid.

Two excellent methods (uiilising acid-ba^e indicators) are avaibble for
Standardisation. Tne lir.st i.s 'widely employed, hut tlic second is more convenient,
less time-consuming, and equally accurate.

;U .Standanlisation with anhydrous sodium carbonate. Piw .xtiJiaw
carhanaic. Analytical reagent-quality sodium carbonate of 99,9 per cent purity
is obtainable commerc-iaUy. This ec'iuains a little moisture and must be
dehydrated by heating at 260 270 C for half an hour and allowed to cool in a
desiccator before iwe. .Alternatively, pure .sixlium c.irhonatc may be prepared by
heating A.R. sodium hydrogenearbonate to 260-270 C for 60-90 minutes: the
temperature must not be .allowed to exceed 270 C. for above this temperature the
sodium carbonate may lose carbon ilioxidc. It has been recommended (Ref. 11)
that the .A.R. sodium hydrogenearbonate be decomposed by adding to hot (85 C)
water, and then tiie hydrated sodium carbemate which crv.stallises out is filtered
off and dehydrated by hciumc. first at 100 C.and finally at 2fO-270 C

In all eases the crucible is allowed to coo! in a desiccator, and, before it i.s quite
cold, the solid is transferred to a warm. dry. glass-stoppcrcd tube or bottle, out of
which, when cold, it may he weighed lapidly as required. It is important to
remember that anhydrous sodium carbonate is hyf.roscopic and exhibits a
tendency to change into the monohydratc.

PmeeJure. Weigh out accurately from a weighinu bottle about 0.2 g of the
pure sodium carbonate into a 250-cm ' conical flask { 1 k'dissolvc it in 50—75 cm'*
of water, and add 2 drops of methyl orange indicator (2) or preferably of methyl
orange-indigo carmine indicator (Section X. 9), whicli gives a very much more
satisfactory end-point.* Rin.se a clean burette three times with 5-cm^ portions of

Thisindic-ator r, prep.-, red In dMotvinc I ppfRieihvIorjncciwd 2.5? of purirVd indigo carmine ini
litre o, distilfed m xtcr. and filtering ihc solution The cotca: change on p-v^ing from alkaline to adJ
ioiution IS from green to macenia nxh a ncuirat-grcv coknir at pit of .about 4

298

' TITRIMETRIC ANALYSIS X, 37

the acid; fill the burette to a point 2-3 cm above the zero mark and open the
stopcock momentarily in order to fill the jet wfith liquid. Examine the jet to see
that no air bubbles are enclosed. If there are, more liquid must be run out until the
jet is completely filled. Re-fill, if necessary, to bring the level above the zero mark;
then slowly run out the liquid until the level is between the 0.0 and 0.5-cm^ marks.
Read the position of the meniscus to 0.01 cm^ (Section HI, 17). Place the conical
flask containing the sodium carbonate solution upon a piece of unglazed white
paper or a white tile beneath the burette,' and run in the acid slowly from the
burette. During the additioii of the acid, the flask must be constantly rotated with
.one' hand whilst the other hand controls the stopcock. Continue the addition
until the methyl orange becomes a very faint yellow or the green colour
commences to become paler, when the methyl orange-indigo carmine indicator is
used. Wash the walls of the flask down with a little distilled water from a wash
bottle, and continue the titration very carefully by adding the acid dropwise until
the colour of the methyl orange becomes orange or a faint pink, or the colour of
the mixed indicator is a neutral grey. This marks the end point of the titration,
and the burette-reading should be taken and recorded in a note-book. The
procedure is repeated with two or three other portions of sodium carbonate. The
first (or preliminary) titration will indicate the location of the true end j)oint
within 0.5 cm^. With experience and care, subsequent titrations can be carried put
very accurately, and should yield concordant results. From the weights of sodium
carbonate and the volumes of hydrochloric acid employed, the strength of the
acid may be computed for each titration. The arithmetical mean is used to
calculate the strength of the solution.

Notes. 1. For elementary students, an approximately O.IN solution of
sodium carbonate may be prepared by weighing out accurately about 1.3 g of
pure sodium carbonate in a weighing bottle or in a small beaker, transferring it to
a 250-cm^ graduated flask, dissolving it in water (Section IH, 32), and making up
to the mark. The flask is well shaken, then 25.00 cm^ portions are withdrawn with
a pipette and titrated with the acid as described above. Individual titrations
should not differ by more than 0. 1 cm^. Record the results as in Section X, 42.

2. To obtain the most accurate results, a comparison solution, saturated with
carbon dioxide and containing the same concentration of sodium chloride (the
colour of methyl orange in a saturated aqueous solution of carbon dioxide is
sensitive to the concentration of sodium chloride) and indicator as the titrated
solution at the end-point, should be used.

The mixed indicator, bromocresol green-dimethyl yellow, may be used with
advantage. The indicator consists of 4 parts of a 0.2 per cent ethanolic solution of
bromocresol green and 1 part of a 0.2 per cent ethanolic solution of dimethyl
yellow; about 8 drops are used for 100 cm^ of solution. The colour change is from
blue to greenish yellow at pH = 4.0-4. 1 ; the colour is yellow at pH = 3.9.

Methyl red may be used as an Indicator provided the carbon dioxide in solution
is expelled at the end point by boiling. This indicator gives a red colour with high
concentrations of carbon dioxide such as ■ are produced during titrations
involving carbonates. Add the standard hydrochloric acid to the cold sodium
carbonate solution containing 3 drops of 0.1 per cent methyl red until the
indicator changes colour. Boil the solution gently (preferably with a small funnel
in the mouth of the flask) for 2 minutes to expel carbon dioxide: the original
colour of the indicator will return. Repeat the process until the colour no longer
changes on boiling. Generally, boiling and cooling must be repeated twice. Care

299

X, 37 QUANTITATIVE INORGANIC ANALYSIS

must be taken to avoid loss of liquid by spattering during boiling. Tlic colour
chanuc is more easily perceived than with met by 1 orange.

Calculation of’ normality. The normality may be computed from the
equation:

NajCO., + 2Ha 2NaCl + CO,-f H;0

but the best method is to derive the normality entirely in terms of the primary
standard substance, here, sodium carbonate. The equivalent (Section X, 3) of
sodium carbonate is 52.9i>7 or 5.3.(10 g. If ibc weight of the sodium carbonate is
divided by (he numher of cm^ of hydrochloric acid to which it is equivalent, as
found by titration, we have the weight of primary standard equivalent to 1 cm^ of
the acid. Thus if 0.2.500 g of sodium carbonate is required for the neutralisation of
4.5.00 cm'’ of hydrochloric acid. 1 cm’ of the acid would be equivalent to 0.2500;
45.00= 0.005556 g of sodium c.irbonaic. The milli-equivalent or the weight in 1
cm’ of A’-sodiuni carbonate solution is 0.053(X) g. Hence the normality of the acid
is 0.005556/0.05.300 = 0. 1 ahS.V.

Another method is the following. 0.2500 g of sodiunt carbonate required 45.00
cm’ of acid, hence I dm’ of acid is equivalent to KXX) >'-0.2500/45.00 = 5.556 g of
sodium carbonate. Hut a dm’ of .V-acid is equivalent to 53.CK) g of sodium
carbonate, hence the acid is 5.556/53.00-= 0.I04S;V.

In the method described in Note 1 above, the normality of the sodium
carbonate is fir.st computed from the weight of sodiunt carbonate used. The mode
of calculation described in Section X. 4 is employed. If 1 is the %'olumc in cm’ of
thcslandard .solution of normality m, required to react completely with l),cm’of
the unknown .solution of normality /%. then;

l = Vi, X Jiff

from whicli the value of n„ is readily deduced. Thus if 1.3890 g of anhydrous
sodium carbonate is dissolved in 250 cm’ of water, the normality of the smlium
carbonate solution is 1.3890 v. 4/53 00= 0.1CI48.V. If 25 cm’ of the .sodium
carbonate solution c.sactly neutralise 25.45 cm’ of the hydrochloric acid, then:

25.00 X 01048 = 25.45 xn„
or the acid is0.1030A'.

B. Standnrdis.ation ag-ainst sodium tctralwratc. The adr.inlages of sodium
tetraborate dccaliydrate (borax) arc; (1) it has a large equivalent. 190.72 g (that of
anhy’drou.s sodium carbonate i.s 53.(K)):(ii) it is ea.sily and economically purified
by rccryslallisaiion: (iii) heating to con'^tant wciglit is not required; (iv) it is
practically non-hygroscopic; and (v) u sharp end-point can be obtained with
methyl red at room tcmi'K.’ratures. since this indicator is not affected by the very
weak boric acid.

^40,’“ +2H" +511,0 = 4H_,nO,

Pure sodium tetraborate. The ,-\.R. salt is recrystallLscd from distilled water:
50 cm- of water js used for every 15 g of solid. Care must be taken that the
crystallisation doc.s not lake place above 55 C; above this temperature there is a
possibility of the formation of the pentahvdrate. since the transition temperature,
decahydratc pentahydrate. is 61 C.'TIic crystals arc filtered at the pump,
washed twice with water, then twice with portions of 95 per cent ethanol. followed
by two portions of A.R. diethyl ether. Fivc-cin’ portions of water, ethanol, and
ether arc used for 10 g of crystals. Each wasliing must be followed by .suction to

300

TITRIMETRIC ANALYSIS X, 38

remove the wash liquid. After the final washing, the solid is spread in a thin layer
on a watch or clock glass and allqwed to stand at room temperature for 12-18
hours. The sodium tetraborate is then dry, and may be kept in a well-stoppered
tube for three to four weeks without appreciable change. An alternative method
of drying is to place the recrystallised product (after having been washed twice
with water) in a desiccator over a solution saturated with respect to sugar
(sucrose) and sodium chloride. The substance is dry after about three days, and
may be kept indefinitely in the desiccator without change. The latter method is
more time-consuming; the product is identical with that obtained by the ethanol-
ether process.

Procedure. Weigh out accurately from a weighing bottle 0.4-0.5 g of pure
sodium tetraborate into a 250-cm^ conical flask (1), dissolve it in about 50 cm^ of
water and add a few drops of methyl red. Titrate with the hydrochloric aicid
contained in a burette (for details, see under A) until the colour changes to pink
(2). Repeat the titration with two other portions. Calculate the strength of the
hydrochloric acid from the weight of sodium tetraborate and the volume of acid
used. The variation of these results should not exceed 1-2 parts per thousand. If it
is greater, further titrations must be performed until the variation falls within
these limits. The arithmetical mean is used to calculate the concentation of the
solution.

Notes. 1. For elementary students, an approximately O.liV solution of
sodium tetraborate may be prepared by weighing out accurately 4.7-4.8 g of A;R.
material on a watch glass or in a small beaker, transferring it to a 250-cm^
graduated flask, dissolving it in water (Section HI, 32), and making up to the
mark. The contents of the flask are well mixed by shaking. Twenty-five cm^
portions are withdrawn with a pipette and titrated with the acid as detailed under
Method A. Individual titration^ should not differ by more than 0.1 cm^.

2. For work of the highest precision a comparison solution or colour standard
may be prepared for detecting the equivalence point. For O.liV solutions, this is
made by adding 5 drops of methyl red to a solution containing 1.0 g of sodium
chloride and 2.2 g of boric acid in 500 cm^ of water; the solution must be boiled to
remove any carbon dioxide which may be present in the water. It is assumed that
20 cm^ of wash water are used in the titration.

Calculation of the normality. This is carried out as described in Method A.
The equivalent of sodium tetraborate is 190.72 g.

C. Standardisation by an iodometric method. The experimental details are
given in Section X, 130.

D. An argentimetric method is described in Section X, 84.

X, 38. PREPARATION OF STANDARD ALKALI. Discussion. The
hydroxides of sodium, potassium, and barium are generally employed for the
preparation of solutions of standard alkalis; they are water-soluble strong bases.
Solutions made from aqueous ammonia are undesirable, because they tend to
lose ammonia, especially if the concentration exceeds 0.5M; moreover, it is a
weak base, and difficulties arise in titrations with weak acids (compare Section X,
15). Sodium hydroxide is most commonly used because of its cheapness. None of
these solid hydroxides can be obtained pure, so that a standard solution cannot
be prepared by dissolving a known weight in a definite volume of water. Both
sodium and potassium hydroxides are extremely hygroscopic; a certain amount
ot alkali carbonate and water are always present. Exact results cannot be

301

X. 38 QUANTITATIVE INORGANIC ANALYSIS

obtained in the presence of carbonate with some indiciitors, and it is therefore
necessary to discuss methods for the preparation of carbonatc-frcc alkali
solutions. For many purposes A.R. .sodium hydroxide (which contains l~2pcr
cent of sodium carbonate) is sufficiently pure.

To prepare carhonafe-frcc sodium hydroxide .solution one of several methods
may be used:

1. Rinse sodium hydroxide sticks rapidly with water; this remove.? the
carbonate from the surface. A solution prepared from the washed sticks is
satisfactory for most purposes.

2. If a concentrated solution of sodium hydroxide (equal weights of sticks or
pellets and w.atcr) is prepared, covered, and allowed to stand, the carbonate
remains insoluble; the clear supernatant liquid may be poured or siphoned off.
and suitably diluted. (Potassium cirbonaic is too soluble in the concentrated
alkali for this method to be applicable.)

3. A method, which yields a product completely free from carbonate ions,
consists in the electrolysis of a saturated solution of A.R. sodium chloride witha
mercury cathode and a platinum anode in the apparatus shown in Fig. X, 14.
About "20-30 enr' of n'-tlistillcd mercury arc placed in a 250-cm'' pear-shaped
Pyrex separatory funnel; 100 -125 cm^ of an almost saturated solution of A.R.
sodium chloride arc then carefully introduced. Two .sliort lengths of platinum
wire arc sealed into Pyrex glass tubing; one of these dip.s Into the meremy
(cathode), and the other into the salt solution (anode). A little mercury is plaa'd

in the glass tubes, and electrical contact is made by means of
amalgamated copper wires dipping into the mercury in the
tubes. Elect roly.'.Ls is carried out using 6-8 volts and 0.5-1
amp for several hours; the funnel is shaken at intcrx’als
in order to break up the amalgam crystals that form on the
surface of the mercury. Tlic weight of the sodium dissolved in
the amalgam may be roughly computed from the total current
passed; the current efficiency i.s 75-80 per cent. When sulfi-
cicnV amalgam has formed, tlw mercury is run into a Pxtcx
flask containing about 100 cm'’ of boiled-out distilled
water and closed with a rubber bung carrying a soda-lime
guard tube. Decomposition of the amalgam, to give the
sodium hydroxide solution, is complete after several days;
after 12-18 hours about 75 per cent of the amalgam is
decomposed.

4. In the anion cx'change method, which is recommended, carbonate maybe
removed from either sodium or potivssium hydroxide. The .solution is passed
through a strong base anion exchange column (c.g.. Zerolit FF or Amberlite
IRA-400) in the chloride form (see Chapter Vll). Initially the alkali hydroxide
converts the resin into the hydroxide form; the carbonate ion has a greater
affinity for the rc.sin than the hydroxide ion, and hence is retained on the resin: the
first portions of the effluent contain chloride ion. If it is desired not to dilute the
standard base appreciably and if chloride ion is objectionable, the effluent is
discarded until it shows no test for chloride. Thus if a column containing one of
the above resins. 35 cm long, is prepared in a 50-cnr' burette, about 150 cm’ of
4 per cent sodium hydroxide solution must be passed through the column of resin
at a flow rate of 5-6 cm’ per minute before the effluent is cbloridc-free;
subsequently the effluent may be collected in a 500 cm’ filter flask with side arm

“Aj

1 ^

Fig. X, 14

302

TITRIMETRIC ANALYSIS X, 38

carrying a soda-lime guard, tube. When about 125 cm^ of liquid have been
collected, 105 cm^ are measured out and diluted with boiled-out-distilled water
to 1 litre. .The resulting sodium hydroxide solution is carbonate-free and is about
O.IM. When the column, becomes saturated with carbonate ion it is readily re-
converted to the chloride form by passing dilute hydrochloric acid through it,
followed by water .to remove the excess acid.

Strong base anion exchangers in the hydroxide form may be used to prepare
standard solutions of sodium or potassium hydroxide using weighed amounts of
pure sodium chloride or potassium chloride. The resin, after conversion into the
hydroxide form by passage of IM-sodium hydroxide (prepared from 18M-
sodium hydroxide so as to be carbonate-free), is washed with freshly boiled
distilled water until the effluent contains no chloride.ions and is neutral; about 2
litres of IM-sodium hydroxide are required for 40 g of resin, and washing is with
about 2 litres of water. About 2.92 g of A.R. sodium chloride, accurately weighed,
are dissolved in 100 cm^ .of water. The solution is passed through the column at
the rate of 4 cm^ per minute; this is followed by about 300 cm^ of freshly boiled
distilled water. The eluate is collected in a 500-cm^ graduated flask by means of
an adapter permitting the use of a soda-lime guard tube. Towards the end the
flow rate is decreased to permit careful adjustment to volume. A ca. O.IM
solution of sodium hydroxide results.

A number of firms supplying laboratory chemicals offer solutions of known
concentration which can be employed for titrimetric analysis, and amongst these
some manufacturers catalogue sodium hydroxide solutions ‘free from carbonate’,
as for example BDH Chemicals Ltd ‘AVS’ range of solutions, and if only
occasional need for carbonate-free sodium hydroxide solutions arises, this is the
simplest way of satisfying this need. The merits of barium hydroxide solution
(Section X, 41) as a carbonate-free alkali should also be borne in mind, but this
suffers from the disadvantage that the maximum concentration available is
between 0.05iV to 0.17V. Whenever carbonate-free alkali is employed, it is
essential that all the water used in the analyses should also be carbon dioxide free.
With de-ionised water, there will be little cause for worry provided that the water
is protected from atmospheric carbon dioxide, and with ordinary distilled wafer,
dissolved carbon dioxide is readily removed by slowly aspirating a current of air
which has been passed through a tube containing soda asbestos or soda lime
through the water for 5-6 hours.

Attention must be directed to the fact that alkaline solutions, particularly if
concentrated, attack glass. They may be preserved, if required, in polythene
bottles, which are resistant to alkali. Furthermore, solutions of the strong bases
absorb carbon dioxide from the air. If such solutions are exposed to the
atmosphere for any appreciable time they become contaminated with carbonate.
This may be prevented by the use of a storage vessel such as is shown in Fig. Ill,
24; the guard tube should be filled with soda-lime or with soda-asbestos. A short
exposure of an alkali hydroxide solution to the air will not, however, introduce
any serious error. If such solutions are quickly transferred to a burette and the
latter fitted with a soda-lime guard tube, the error due to contamination by
carbon dioxide may be neglected.

The solution of alkali hydroxide prepared by any of the above methods must be
standardised. Alkaline solutions that are subsequently to be used in the presence
of carbon dioxide or with strong acids are best standardised against solutions
prepared from constant boiling-point hydrochloric acid or potassium hydrogen-

303

X, 39 QUANTITATIVE INORGANIC ANALYSIS

iodatc or sulphamic acid, or againsi hydrochloric acid which has been
Standardised by means of .sodium tetraborate or sodium carbonate. If the alkali
.solution is to be u.scd in the titration of weak acids, it is bc.si .standardised against
organic acids or against acid .salt oforganicdiproiicacid.s, such as benzoic acid or
potassium hydrogenphthalatc. respectively. These .substances arc commercially
available in a purity c.xceeding 99.9 per cent; potassium hydrogenphthalatc is
preferable, since it i.s more .soluble in water and has a greater equivalent.

Procedure A. Weigh out rapidly about 4,2 g of A.R. sodium hydroxide on a
watch glass or into a small beaker, dissolve it in water, make up to 1 litre with
boilcd-out distilled water, mix thoroughly by sitaking. and pour the resultant
solution into the stock bottle, which should he closed by a rubber stopper.

Procedure li (carbonatc-frcc sodium hydroxide). Dissolve 50 g of sodium
hydroxide in 50 cm’ of distilled water in a Py rex flask, transfer to a 75-cm’ test-
tube of Pyrex glass, and insert a well-fitting stopper covered with tinfoil. Allow it
to stand in a vertical position until the supernatant liquid is clear. Fora O.l.M-
sodium hydroxide solution enrefuliy withdraw', using a pipette fitted with a filling
device, fi.5 cm’ of the concentrated clear solution into a litre bottle or fla.sk. and
dilute quickly with 1 litre of recently boilcd-out w ater.

A clear solution can be obtained more quickly, and incidentally the Iramsfcr
can be made more satisfactorily, by rapidly filtering the solution through a
sintered glass funnel with exclusion of carbon dioxide with the aid of the
apparatus shown in Fig. X. 1.5. It is advisable to calibrate the test-tube in
approximately 5-cm-' intervals and to put the graduations on a thin slip of paper
gummed to the outside of the tube.

X,39. STANDARDISATION OF THE APPROXIMATELY O.IM-
SODIUiM ID DROXIDE. If the solution contains carbonate {Procedure A).
methyl orange, methyl orange-indigo carmine, or bromophcnol blue must be
used in stnndardi.s.ation against hydrochloric acid of known normality.
Phcnolphthaicin or indicators with a similar pH range, which are afli-’Cted by
carbon dio,xide, cannot be used at the ordinarv temperature (compare Section X,
7). With carbonate-frcc sodium hydroxide '{Procedure li) phcnolphthaicin or
thymol blue {.Section X, 13) may be employed, and standardisation may be
effecled against hydrochloric acid, potassium hydrogeniodatc. potassium
lijdrogenphthalatc. benzoic acid, or other organic acids (Section X. 40).

Procedure A. With st.andard hydrochloric acid. Place the siandardi.sed
(approx. 0.1 Af) hydrochloric acid in the burette. Transfer 25 cm’ of the sodium
hydroxide solution into a 250-cm’ conical flask with the aid of a pipette, dilute

304

TITRIMETRIC ANALYSIS X, 40

with a little water, add 1-2 drops of methyl orange or 3-4 drops of methyl
orange-indigo carmine indicator, and titrate with the previously standardised
hydrochloric acid. Repeat the titrations until duplicate determinations agree
within 0.05 cm^ of each other.

Calculation of the normality. The normality is readily computed from the
simple relationship; ■

where and refer to the volume and known normality of the acid respectively,
Vg is the volume of alkali solution required for the neutralisation, and tig is its
(unknown) normality.

Procedure B. With potassium hydrogenphthalate. A.R. potassium
hydrogenphthalate has a purity of at least 99.9 per cent; it is almost non-
hygroscopic, but, unless a product of guaranteed purity is purchased, it is
advisable to dry it at 120 °C for 2 hours, and allow it to cool in a covered vessel in
a desiccator. Weigh out three 0.6-0.7 g portions of the salt into 250 cm^ Pyrex
conical flasks (1), add 75 cm^ of boiled-out water to each portion, stopper each
flask and shake gently until the solid has dissolved. Titrate each solution with the
sodium hydroxide solution contained in a burette, using phenolphthalein or
thymol blue as indicator.

Calculation of normality. This is similar to that described in Section X, 37.
The equivalent of potassium hydrogenphthalate is 204.22 g. The variation in the
results should not exceed 0. 1-0.2 per cent.

HK(C 8 H 40 J + NaOH = NaK(C 8 H 404 ) -t- H 2 O

Note. 1. For elementary students, an approximately O.IM solution is
prepared by weighing out accurately about 5.1 g of the ordinary A.R. product,
dissolving it in water, and making it up to 250 cm^ in a graduated flask. Twenty-
five-cm^ portions are employed in the titrations with the sodium hydroxide
solution. Individual titrations should not differ by more than 0.1 cm^.

X, 40. OTHER STANDARD SUBSTANCES FOR AODIMETRY AND
ALKALIMETRY. In addition to the standard substances already detailed for
use in standardising acids and alkalis, numerous others have been proposed. A
number of these will be briefly described.

A. Benzoic acid (CgHjCOOH; equivalent = 122.12 g). The A.R. product
has a purity of at least 99.9 per cent. For work demanding the highest accuracy,
the acid should be dried before use by careful fusion in a platinum crucible placed
m an oven at about 130 °C, and then powdered in an agate mortar. Benzoic acid is
sparingly soluble in water (which is a disadvantage) and must therefore be
dissolved in 95 per cent ethanol. The mode of use is similar to that already
described for potassium hydrogenphthalate (Section X, 39, B). For a O.IM
solution, of, say, sodium hydroxide, weigh out accurately 0.4 g portions of the
acid into a 250-cm^ conical flask, add 10-20 cm^ of ethanol, shake until dissolved,
and then titrate the solution with the strong alkali using phenolphthalein as
indicator. A blank test should be made with the same volume of ethanol and the
indicator; deduct, if necessary, the volume of the alkali solution consumed in the
blank test.

B. Succinic acid {(CH 2 COOH) 2 ; equivalent = 59.045 g). The A.R.
product or a pure commercial product should be recrystallised from pure acetone

305

X, 41/42 QUANTITATIVE INORGANIC ANALYSIS

and dried in a vacuum desiccator. The purity is checked by means of a melting,
point determination (185-185.5 C), The acid is fairly soluble in water;
phcnolphthalcin is a .suitable indicator.

C Potassium hydrogeniodate JK 11 ( 103 ),; equivalent 389.95 g}. Unlike
the other solid standards already described, this is a .strong acid and thus permits
the use of any indicator having a pH range between 4.5 and 9.5 for titration with
strong bases. It may be employed for the .standardisation of weak bases which arc
subsequently to Isc used with strong acids; an indiaitor, such as methyl red.
must then be used. The salt is moderately soluble in water (1.33 gi' 100 cm^ at IS''),
is anhydrous and non-hygroscopic. and its aqueous solution is stable for long
period's: the equivalent is high, and a 0.01 A’ .solution contains 3.8995 g pcrdm\

Preparaiion of pure pmassiinn hyilro'^eniodatc. Dissolve 27 g of A.R,
potassium iodate in 125 cm ' of boiling water, and add a solutioti of 22 g of A.R.
iodic acid in 45 env’ of w'arm water acidified with 6 drops of concentrated
hydrochloric acid. Pota.s,siiim hjdrogcniodate separates on cooling. Filter on a
sintercd-glass funnel, anti wash with cold water. Rccrysiallisc three times from
hot water: use 3 parts of water for I part of the salt ;tnd stir continuously during
each cooling. 13ry the crystals at IfK) C? for several hours. Tlic purity c.xceeds
99.95 per cent.

D. Sulphnniic!Kid(NlljSO;OH:cquivalcnt ~ 97.09g). A product ofhich
purity (>99.9 per cent) is available commercially. It is a colourless, crystalline,
non-hygroscopic solid, melting with decomposition at 205 'C. Tlic acid is
moderately soluble in water (21.3 g and 47.1 gin 100 g of water at 20 and 80 '"C
respectively). Sulphtimic acid acts as a strong acid, so tliat any indictuor with a
colour change in the pH range 4-9 may be emploscd; bromothymol blue is
particularly suitable for use with strong ba.scs. It undergoes hydrolysis in aqueous
solution.

NH,SOjOIH HjO NHJISO^

AqucoiKs solutions should, preferably, not be stoicd; the litre does not alter on
keeping if an indicator which changes in the acid range is used.

X.41. STANDARD BARIILM HVDFmXIDE (BARYTA) SOLU-
TION. This solution is widely employed, particularly for the titration of
organic acids. Barium carbonate is insoluble, so that a clear solution is a
carbonaic-frce .strong alkali. The equivalent orBa(OH),.8H;0 is 157.75 g.buta
standard solution cannot be jucparcd by direct weighing owing to the
uncertainty of the hydration end the po.^sib!e presence of carbonate. To prepare
an appro,\imately O.l.Y solution, dissolve IS g of A.R. crystallised barjta (or 20g
of the commercial .substance) in about I litre of Itol water in a large flask. Stopper
the flask and allow the solution to stand for 2 davs or until all the barium
carbonate has completely settled out. Decant or siphon oflT the clear solution into
a storage bottle of the type depicted in Fig. Ill, 24. A soda-lime guard tube must
be provided to prevent ingrc.ss of carbon dio.vidc. The .solution may be
stand, irdi.sed against .standard 0, l,\/-hydrochIoric acid, succinic acid or
potassium liydrogenphllialatc; phcnolphthalcin or thymol blue is employed as
indicator.

X, 41 DETERMFNATIO.N OF THE NaXO., CONTENT OF WASHING
SODA. Procedure. Weigh out accurately about 3.6 g of (lie washing-soda

306

TITRIMETRIC ANALYSIS X, 42

crystals, dissolve in water, and make up to 250 cm^ in a graduated flask. Mix
thoroughly. Titrate 25 cm^ of the solution with standard hydrochloric acid of
approximately O.IM concentration using methyl orange, or, better, methyl
orange-indigo, carmine or bromo-cresol green as indicator. Two consecutive
titrations should agree within 0.05 cm^

Calculation. The weight of anhydrous sodium carbonate Na2C03 which
has reacted with the standard hydrochloric acid can be readily computed from
the equation:

NajCOj +2HC1 = 2NaCl + H2O + CO2
106.01 2x36.46

The percentage of Na2C03 can then be calculated from the known weight of
washing soda employed.

A simpler and more general procedure is to employ the normality method. An
actual example will make this clear.

Weight of weighing bottle + substance = 16.7910 g

Weight of weighing bottle + residual substance = 13.0110 g

.'. Weight of sample used = 3.7800 g

This was dissolved in water and made up to 250 cm^.

Titration of 25.00 cm^ of the carbonate solution with 0.1060N-HC1, using
methyl orange-indigo carmine as indicator.

Experiment

Reading 1

Reading 2

Difference

0.50 cm^

26.60 cm*

26.10 cm* (preliminary)

0.55 cm^

26.45 cm*

25.90 cm*

0.50 cm^

26.45 cm*

25.95 cm*

Mean 25.93 cm*

1 cm^ M-HCl = 0.05300 g Na2C03

25.93 X 0.1060 s 2.749 cm^ M-HCl

2.749 X 0.05300 = 0.1457 g Na2C03 in portion titrated.

Weight of washing soda in portion titrated

= 3.7800 X 25.0/250 = 0.3780 g

.•. Percentage of Na2C03 = 0.1457 x 100/0.3780 = 38.54 per cent. !
Alternative method of calculation. 25.0 cm^ of the carbonate solution

required 25.93 cm^ of 0.1060M-HC1: . , . ■

.". 25.0 X normality of carbonate solution = 25.93 x 0.1060, whence the
carbonate solution is 25.93 x 0.1060/25.0 = 0.1099N. But JV-Na2C03 contains
106.00/2 = 53.00 g Na2C03 per dm^.

. . the given solution contains 0.1099 x 53.00 = 5.8271 g Na^COj per dm^;

307

X, 43 QUANTITATIVE INORGANIC ANALYSIS

and 250 cnr’ would coniain 5.S27I x 250/1000 ^ 1.4568 g.

TliuspcrccniagcofNaXOj ^ l.4568x 100/3.7800 “ 38.54 gormif.

X. 43. DETERMINATION OFTHESTRENGTH OF CONCENTRATED
ACl ns. la) Glad.al acclic ncld. Weigh a dry, stoppered 50 enr' conical fla.sk,
introduce about 2 g of glacial acetic acid and weigh again. Add about 20 cm^ of
water and transfer the" solution quantitatively to a 250-cm-’ graduated flask.
Wash the small flask scserai fime,s' with water and add the washings to the
graduated flask. Make up to the mark with distilled, prefer.ibly boiled-out stater.
Shake the flask well to ensure thorough mixing. Titrate 25-crn-' portions of the
acid with 0.1 A/ standard sodium hydroxide solution, using plicnolphthalcin or
thymol blue as iiidicritor.

NaOII -t-CHjCOOH - CH,COONa + 11. 0
1 cm-’ A/-NaOIl r 0.06005 g Cll.,COOH

Calculate the perccntiigc of Cl 1 jCOOH in the sample of gJacia! acetic acid.

Note on the determination of the acetic acid content of vinegar, \flnegar
usually contains 4 5 per cent acetic acid. Weigh out about 20 g vinegar as
described above, and make up to 100 cm’ in a graduated flask. Remose 25 cm’
with a pipette, dilute with an equal volume of water, add a few drops of
plicnolphthalcin, and titrate with standard 0.1.\/-sodium hydroxide solution.
As a result of the dilution of the vinegar, its natural colour will be so reduced that
it will not interfere witli the colour change of the indicator. Calculate the acetic
acid content of the vinegar, and express the result in p of acclic acid per 100 grams.

(b) Concentrated sulphuric acid. Place about 10f}cm’ofwntcrina250cni’
graduated flask, and insert a short-stemmed funnel in the neck of the flask.
Charge a weight pipette with a few grams of the acid to be evaluated, and weigh.
Add about 1. 3-1. 5 g of the acid to the flask and reweiph the pipette. Alternatively,
the acid may be weighed out in a stoppered weighing bottle, and after adding the
acid to the flas’k, the weighing bottle is reweighed. Rinse llic funnel thoroughly,
remove the flask, and allow the flask to stand for 1-2 hours to regain the
temperature of the laboratory when the solution can be made up to the mark.
Shake and mix thoroughly, and then titrate 25 cm’ portions with standard 0.1 Af
sodium hydroxide, using methyl orange or methyl orange -indigo carmine a,s
indicator.

1 cm’ A/-NaOH re 0.4904 g H.SUj.

Fuming sulphuric acid (oleum) should be weighed in a Lungc-Rey pipette
(Fig. Ill, n,b).

(c) Syrupy phosphoric ns- d. In tins ease we arc dealing with a triproticacid
and theoretically three equivalence points are possible, but in practice the pH
changes in the neighbourhood of the equivalence points are not very marked (see
Fig. X. 6), For the first stage neutralisation (pH 4,6) we may employ methyl
orange, methyl orange -indigo carmine or bromocresol green as indicator, but it

to use a comparator solution of sodium diliydrogcnphosphalc
(0.03AI) containing the same amount of indicator as in the solution being titrated,
ror the .second stage (pH 9.7), phcnolplulialein is not altogether .salksfactoiy (it
changes colour on the early side), tlivmolpluhalcin is belter; but the best
indiaitor is a mixture of phcnolphthalcin (2 parts) with l-naphtliolplithalein

308

. TITRIMETRIC ANALYSIS X, 44

(1 part) which changes from pale rose through green to violet at pH 9.6. For the
third stage (pH 12.6) there is no satisfactory indicator.

Procedure. Weigh an empty stoppered weighing bottle, add about 2 g of
syrupy phosphoric acid and reweigh. Transfer the acid quantitatively to a 250
cm^ graduated flask, and then proceed as detailed for sulphuric acid, but using
the phenolphthalein- 1-naphtholphthalein mixed indicator. .

H 3 P 04 + 20 H“ = HP0^'*"+2H20 ,

1 cm^ M-NaOH = 0.04902 g H3PO4

X, 44. DETERMINATION OF A MIXTURE OF CARBONATE AND
HYDROXIDE. Analysis ofcommercial caustic soda. Discussion. Two methods
may be used for this analysis. In the first method the total alkali (carbonate +
hydroxide) is determined by titration with standard acid, using methyl orange,
methyl orange-indigo carmine, or bromo-phenol blue as indicator. In a second
portion of solution the carbonate is precipitated with a slight excess of barium
chloride solution, and, without filtering, the solution is titrated with standard
acid using thymol blue or phenolphthalein as indicator. The latter titration gives
the hydroxide content, and by subtracting this from the first titration, the volume
of acid required for the carbonate is obtained.

Na^COj-t-BaClj = BaC03 (insoluble) -i-2NaCl

The second method utilises two indicators. It has been stated in Section X, 17
that the pH of half-neutralised sodium carbonate, i.e., at the sodium
hydrogencarbonate stage, is about 8.3, but the pH changes comparatively slowly
in the neighbourhood of the equivalence point; consequently the indicator
colour-change with phenolphthalein (pH range 8.3-10.0) or thymol blue (pH
range (base) 8.0-9.6) is not too sharp. This difficulty may be surmounted by using
a comparison solution containing sodium hydrogencarbonate of approximately
the same concentration as the unknown and the same volume of indicator. A
simpler method is to employ a mixed indicator (Section X, 9) composed of 6 parts
of thymol blue and 1 part of cresol red; this mixture is violet at pH 8.4, blue at pH
8.3, and rose at pH 8.2. With this mixed indicator the mixture has a violet colour
m alkaline solution and changes to blue in the vicinity of the equivalence point; in
making the titration the acid is added slowly until the solution assumes a rose
colour. At this stage all the hydroxide has been neutralised and. the carbonate
converted into hydrogencarbonate. Let the volume of standard acid consumed be
rcm^.

0 H-+H+=H20

C03^-+h+ =hco3- ;

Another titration is performed with methyl orange, methyl orange-indigo
carmine or bromophenol blue as indicator. Let the volume of acid be V cm^.

OH--f-H+ = H20

C03^--f2H+ =H2C03

H2C03^H20-fC02

Then V~2(V-v} corresponds to the hydroxide, 2{V-v) to the carbonate, and F to

309

X, 45 QUANTITATIVE INORGANIC ANALYSIS

the total alkali. To obtain satisfactory results by this method the .solution titrated
must be cold (as near 0 C as i.s practicable), and loss of carbon dioxide mu,st be
prevented as far a.s possible by keeping the tip of the burette immersed in the
liquid.

Procedure A. Weigh out accurately in a glass-stoppered weighing bottle
about 2.5 g of commercial sodium hydroxide (e.g.. in flake form). Transfer
quantitatively to a 500-cm^ graduated Das): and make up to the mark. Shake the
lliusk well Titrate 25 or 5(1 cm-' of this solution with standard O.l .M-hydrochloric
acid, using methyl orange or methyl orange- indigo carmine as indicator. Carry
out two or three titrations; these should not differ by more than O.I cm'. This
gives the tot.al alkalinity (hydroxide -i carbonate). Warm another 25 or 50 cm'
portion of the solution to 70 C and add I per cent barium chloride solution
slowly from a burette or pipette in .di.g/if excess, i.c., until no further precipitate i.s
produced. Cool to room temperature, add a few drops of phcnolphthalcin to the
.solution, and titrate very .slowly and with constant stirring with standard 0.!.\f-
hydrochloric acid; the end-point is reached when the colour ju.st changes from
pink to colourless. If thymol blue is used as indicator, the colour change is from
blue to yellow. Tlic amount of acid used corresponds to the hydro.xidc present.

This method yields only ;ipproximatc rc.suUs because of the precipitation of
basic barium carbonate in the presence of hydroxide. More accurate results are
obtained by considering the above titration ns a preliminary one in order to
ascertain the approximate hydroxide content, and then carrying out another
titration as follows. Treat 25 50 cm’ of the .solution with .sufTicient standard
hydrochloric acid to nculralisc most of the hydroxide, then heat and precipitate
as before. Under these conditions, practically puie barium carbonate is
precipitated.

1 cm' .M-HCl r: 0.040! p NaOH

1 cm' Af-MCl 2 0.053CK)g Na.CO,

Procedure B. The experimental details for tlic preparation of the initial
solution arc similar to those given under Procedure A. Titrate 25 or 50 cm' of the
cold solution with standarri 0.1. (/-hydrochloric ;irid and methyl orange, methyl
orange -indigo carmine, or bromo-plsenol blue as indicator. Titrate another 25 or
50 cm' of the cold solution, diluted 'v.ih an equal volume of w.ater. slowly with the
standard acid using phcnolphtl..... ir. better, the tltymol bluc-cresol red mixed
indicator; in the latter case, the colour at the end-point is rose.

Calculate the result as described in the /fi.sfiovmn above.

X, 45. DETERMINATION OF A MIXTURE OF CARBONATE AND
HA DROGENCARBON.ATf'l The two methods av:iilable for this de-
termination arc modifications of those described in the previous .Section for
hydroxide-carbonate mixtures. In lire fust procedure, which is particularly
valuable when the sample contains relatively large amounts of c-arbonatc and
small amounts of hydrogencarbonatc. the total afkidi is fir.st determined in one
portion of the solution by titration with standard 0,1 Af-hydrochloric acid using
metliyi orange, methyl orange- indico carmine, or bromophcnol blue as
indicator;

COj'--t-2ir = HXOj

310

TITRIMETRIC ANALYSIS X, 46

HCO 3 -+H+ =H2C03

H2C03^H20 + C02

Let this volume correspond to V cm^ M-HCl. To another sample, a measured
excess of standard O.lM-sodium hydroxide (free from carbonate) over that
required to transform the hydrogencarbonate to carbonate is added;

HCO3 - + OH - = CO3'' - + HjO

A slight excess of 10 per cent barium chloride solution is added to the hot solution
to precipitate the carbonate as barium carbonate, and the excess of sodium
hydroxide solution immediately determined without filtering off the precipitate
by titration with the same standard acid ; phenolphthalein or thymol blue is used
as indicator. If the volume of excess of sodium hydroxide solution added be
equivalent to v cm^ of M-sodium hydroxide and v' cm^ M-acid corresponds to
the excess of the latter, then v-v’ = hydrogencarbonate, and V—(v-v’)
= carbonate.

In the second procedure a portion of the cold solution is slowly titrated with
standard 0.1 M-hydrochl otic acid, using phenolphthalein or, better, the thyiriol
blue-cresol red mixed indicator. This (say, Y cm^) corresponds to half the
carbonate (compare Section X, 44):

COj^'+H+^HCOj-

Another sample of equal volume is then titrated with the same standard acid
using methyl orange, methyl orange-indigo carmine or bromophenol blue as
indicator. The volume of acid used (say, y cm^) corresponds to carbonate +
hydrogencarbonate. Hence 27 = carbonate, and y- 27 = hydrogencarbonate.

X,46. DETERMINATION OF BORIC ACID. Discussion. Boric acid acts as
a weak monoprotic acid (X„ = 6.4 x 10“*“); it cannot therefore be titrated
accurately with 0.1 JV standard alkali (compare Section X, 13). However, by the
addition of certain organic polyhydroxy compounds, such as mannitol, glucose,
sorbitol, or glycerol, it acts as a much stronger acid (for mannitol 1.5 x lO""^)

and can be titrated to a phenolphthalein end-point.

The effect of polyhydroxy compounds has been explained on the basis of the
formation of 1,1- and 1,2-mole ratio complexes between the hydrated borate ion
and 1,2 or 1,3 diols:

>qOH)

2 +H 3 BO 3 =

>C(OH)

>c— o o— c<

\ /

/ \

>c— o o— c<

H++3H,0

Glycerol has been widely employed for this purpose but mannitol. and sorbitol
are more effective, and have the advantage that being solids they do not
materially increase the volume of the solution being titrated: 0.5-0.7 g of
mannitol or sorbitol in 10 cm^ of solution is a convenient quantity.

The method may be applied to commercial boric acid, but as this material may
contain ammonium salts it is necessary to add a slight excess of sodium carbonate
solution and then to boil down to half bulk to expel ammonia. Any precipitate
which separates is filtered off and washed thoroughly, then the filtrate is

311

X, 47 QUANTITATIVE INORGANIC ANALYSIS

neutralised to methyl red, and after boiling, mannitol is added, and the solution
titrated with standard 0.1 Af-sodium hydroxide .solution;

Hlboric acid complex] + NaOH - Na[boric acid complex] + HjO
1 cm^ .A/-NaOH = 0.06184 g

A mixture of boric acid and a strong acid can be analysed by first titrating the
strong acid using methyl red indicator, and then after adding mannitol or
sorbitol, the titration is continued using pbcnolpbihalein as indicator. Mixtures
of sodium tetraborate .and boric acid can be similarly tinalyscd by titrating the
salt reith standard h\drodik)ricacid (Section X. .^7, B), and then adding mannitol
and continuing the titration with standard sodium hydroxide solution; it must of
course be borne in mind that in this second titration the boric acid liberated in
the first titration will also react.

J'rorrdurc. To determine the purity of a sample of boric acid, weigh
accurately about 0.8 g of the acid, transfer quantitatis cly to a 2.80 an-' graduated
flask and make up to the mark. Pipette 2.“' cm' of the solution into a 250 cm'
conical flask, .-idd an equal volume of distilled w.iicr, 2.5~.A g of mannitol or
sorbitol, and titrate with standard O.I.\f-sodium hydro.xide solution using
phcnolphthalcin as indicator. It is advisable to check whether any bkank
correction must be made: dissohe a similar weiglit of tnannitol (sorbiloli in 50
cm' distilled water, add pheiu'lphtbalein. and ascertain how much sodium
hydroxide solution must he added to produce the characteristic end-point colour.

X,47. DF.TKRMINATIO.N OK AM.MO.NIA IN AN A.M.MO.MUM
S.ALT, Disacision. Tu o methods, the direct and indirect, may be used for this
determination. In the direct method, a solution of the ammonium salt is tre,ated
with a solution of a strong b;i.sc(e,g.. .sodium hydroxide) and the mixture distilled.
Ammonia is quantnaiixely cxpcllcvl. and i.s absorlvd in an excess of standard
acid. The excess of acid is buck-tiir.itcd m the presence of methyl red (or methyl
orange, methyl orange- indigo caiminc. bronio-phcnol blue, or bromo-crcsol
green). Hach cm' of ,Y acid consumed in the reaction is equivalent to 0,0170.52 g
NHj:

NH.'+OIl — NtljIflKO

In the indirect method, the aninumium salt (other than the carbori.ite or
bicarbonate) is boiled with a known excess of standard sodium hydroxide
solution. TIic boiling is continued iit.,ii no morcammoniti escapes with tlic.sicarn.
The excess of sodium hydroxide is titrated with standard acid, using methyl red
for methyl orange -indigo c;irminc)as iiuiicator.

Procedure (direct method). Fit up the apparatus shown in Fig. X. 16. note
that in order to provide some flcvibiliiy. ihe spray trap isjoined to the condenser
by a hemispherical ground joint and this makes it easier to clamp both the flask
and the condenser without introducing anv strain into the assembly. The flask
may be of round bottom form (capiidiy 5tX)-1000 cm'), or {as shown in the
di.igram), a Kjcldahl flask. The latter is particularly suitable when nitrogen in
organic compounds is determined by the Kjcldahl method: upon completion of
the digestion with concentrated sulphuric acid, coolini’. and dilution of Ihe
contents, the digestion (Kjcldahl) flask is attached to the apparatus as shown in
Fig. X. 16. The purpose of the spray trap i.s to prevent droplets of sodium

312

TITRIMETRIC ANALYSIS X, 47

hydroxide solution being driven over
during the distillation process. The
lower end of the condenser is allowed
to dip into • a known volume of
standard acid contained in a suitable
receiver, e.g. a conical flask. A com-
mercial distillation assembly may be
purchased (Quickfit and Quartz Ltd),
in which the tap funnel shown is
replaced by a special liquid addition
unit : this is similar in form to the tap
funnel, but the tap and barrel are
replaced by a small vertical ground
joint which can be closed with a
tapered glass rod. This modification
is especially useful when numerous
determinations have to be made as it
obviates the tendency of glass taps to
‘stick’ after prolonged contact with
concentrated solutions of sodium
hydroxide.

For practice, weigh out accurately
about 1.5 g of A.R. ammonium
chloride, dissolve it in water, and
make up to 250 cm^ in a
graduated flask. Shake thoroughly.

■ Transfer 50.0 cm^ of the solution into the distillation flask and dilute with 200
cm^ of water: add a few anti-bumping granules (fused alumina) to promote
regular ebullition in the subsequent distillation. Place 100.0 cm^ of standard 0.1
M-hydrochloric acid in the receiver and adjust the flask so that the end of the
condenser just dips into the acid. Make sure that all the joints are fitting tightly.
Place 100 cm^ of 10 per cent sodium hydroxide solution in the funnel. Run the
sodium hydroxide solution into the flask by opening the tap ; close the tap as soon
as the alkali has entered. Heat the flask so that the contents boil gently. Continue
the distillation for 30-40 minutes, by which time all the ammonia should have
passed over into the receiver ; open the tap before removing the flame. Disconnect
the trap from the top of the condenser. Lower the receiver and rinse the condenser
with a little water. Add a few drops of methyl red* and titrate the excess of acid in
the solution with standard 0.1 M-sodium hydroxide. Repeat the determination.

Calculate the percentage of NH 3 in the solid ammonium salt employed.

1 cm^ O.lAf-HCl = 1.703 mg NH 3

Procedure (indirect method). Weigh out accurately 0.1-0.2 g of the
ammonium salt into a 500-cm^ Pyrex conical flask, and add 100 cm^ of standard
0.1 M-sodium hydroxide. Place a small funnel in the neck of the flask in order to

A sharper colour change is obtained with the mixed indicator methyl red-bromo-cresol green
(prepared from 1 part of 0.2 per cent methyl red in ethanol and 3 parts of 0. 1 per cent bromo-cresol
green in ethanol). ‘ . ■ ■

313

X, 48/49 QUANTITATIVE INORGANIC ANALYSTS

prevent mechaniail loss, and boil the mixture until a piece of filter paper
moistened with mcrcuryfl) nitrate .solution and held in the escaping steam is no
longer turned black. Cool the .solution, add ;i few drops of methyl red. and titrate
with standard 0. 1 A/diydrochloric acid. Repeat the determination.

X. 48. DCTERMINATION OF MTRATliS. Dhaissian. Nitrates arc redured
to ammonia by means of aluminium, zinc or, most conveniently, by Devarda's
alloy (50'>„Cu, 45" ;,.M, 5't-;,Znl in strongly alkaline solution-.

3NO.r »-SAl + 50ir +2H,0 SAIOC r3N!lj

The ammonia is distilled into excess of standard acid as in the prcviou.s Section.

Nitrites are .similarly reduced, and must be allowed for if nitrate alone is to be
determined.

Procedure. Weigh out .accurately about 1.0 g of the nitrate. Dissolve it in
water and transfer the solution quantitatively to the distillation flask of F'ig. X, 16.
Dilute to about 240 cm'. Add 3 g of pure, finely di\ ided Devarda’s alloy fit should
all pass a ZO-mcsIi siese). I'it up tlie apparatus completely and place 75-100 cm^
standard 0.2,V-hydiochloric acid in the receiver fSOO-cm' Pyre.x conical flask)
Introduce 10 cm * of 20 per cent sodium hydroxide solution through the funnel,
and immcdiutch close the tap. Warm genrfy to start the reaction, and allow the
apparatu.s to stand for an hour, by -.vhich time the evolution of hydrogen should
have practically ceased ;md the reduction of nitrate to ammonia be complete.
Then boil the liquid gently and continue the distillaiion until 40-50 cm-’ of liquid
remain in the distillation llask. Open the tap before removing ilie flame. Wash the
condenser ss itli a little distilled water, and titrate the content.s of the rcanver plus
the washings witli standard 0 . 2 Af-sodium hydroxide, using methyl red as
indicator. Rcfscat the determination For very accurate work, it is recommended
that a blank test be carried (uii whIi distilled water,

1 cm’ Af-HC! r. 0.06201 g NOj

X,49. ni'TF.RMlNATION OF I’HOSPHATi: {PRECIPIT.ATION AS
QUINOLINE MOLVBI)OPHOSPn..\TE). ■w^skm. When a .solution of
an orthophosphate is treated with a large excess of ammonium molybdate
.solution in the presence ofnilncacid at a temperature of 20 -45 '^C.:i precipitate is
obtained, which after washing is converted into ammonium molvbdophosphate
with the composition (NH 4 )j 1 PO..i 2 M 0 O 3 ]. Tliis may be titrated with standard
.sodium h 3 ’droxide solution using phenolphllialcin as indicator, but the end-point
is rather poor due to the liberation of ammonia. If, however, the ammonium
molybd.-ite is replaced by a reagent containing sodium molybdate and quinoline,
then quinoline molybdopho.sphatc is prccijiitatcd which can be isolated and
litnited witli standard .sodium h.vdroxide:

(C,H,N),fP04.I2MoO,)-f-26NaOH

- NajHPO^ + 12Na,Mo04 -!-3C„n.N + 14fl,0

The main advantages over the ammonium molybdopho-sphate method arc: (i)
quinoline molybdopliosphatc is Ic.ss soluble and has a constant composition, and
(n) quinoline is a sullicicnily weak base not to interfere in the titration.

C.ilcium, iron, magnesium, alkali metals, and citrates do not alTcci ilicanalysis.
Ammonium salt.s interfere and must be eliminated by means ofsodiiim nitrite or

314

TITRIMETRIC ANALYSIS X, 49

sodium hypobromite. Tlie hydrochloric acid normally used in the analysis may
be replaced by an equivalent amount of nitric acid without any influence on the
course of the reaction. Sulphuric acid leads to high and erratic results. If
hydrochloric acid is present in amount slightly in excess of the sulphuric acid the
interference is prevented, but the total acidity should not be greatly in excess of
2N.

The method may be standardised, if desired, with pure potassium dihydrbgen
phosphate (see below): sufficient i:l-hydrochloric acid must be present to
prevent precipitation of quinoline molybdate; the molybdophosphate complex is
readily formed at a concentration of 20 cm^ of concentrated hydrochloric acid
per 100 cm^ of solution, especially when warm, and precipitation of the quinoline
salt should take place slowly from boiling solution. A ‘blank’ determination
should always be made; it is mostly due to silica.

Solutions required. Sodium molybdate solution. Prepare a 15 per cent
solution of A.R. sodium molybdate, Na 2 Mo 04 , 2 H 20 . Store in a polythene
bottle.

Quinoline hydrochloride solution. Add 20 cm^ of redistilled quinoline to 800
cm^ of hot water containing 25 cm^ of pure concentrated hydrochloric acid, and
stir well. Cool to room temperature, add a little filter paper pulp (‘accelerator’),
and again stir well. Filter with suction through a paper-pulp pad, but do not
wash. Dilute to 1 litre with water.

Mixed indicator solution. Mix two volumes of 0.1 per cent phenolphthalein
solution and three volumes of 0.1 per cent thymol blue solution (both in ethanol).

To standardise the procedure, A.R. potassium dihydrogenphosphate which
has been dried at 105 °C is usually suitable; if necessary it may be further purified
by dissolving 100 g in 200 cm^ of boiling distilled water, keeping on a boiling
water bath for several hours, filtering through paper pulp from any turbidity
which may appear, and cooling rapidly with constant stirring. The crystals are
filtered with suction on hardened filter paper, washed twice with ice-cold water
and once with 50 per cent ethanol, and dried at 105 °G.

Procedure. This will be described by reference to the standardisation with
potassium dihydrogenphosphate. Weigh accurately 0.08-0.09 g of the pure salt
into a 250 cm^ conical flask and dissolve in about 50 cm^ of distilled water. Add
20 cm^ of concentrated hydrochloric acid, then 30 cm^ of the sodium molybdate
solution. Heat to boiling, and add a few drops of the quinoline reagent from a
burette while swirling the solution in the flask. Again heat to boiling arid add the
quinoline reagent drop by drop with constant swirling until 1 or 2 cm^ have been
added. Boil again, and to the gently boiling solution add the reagent a few cm^ at
a time, with swirling, until 60 cm^ in all have been introduced. A coarsely
crystalline precipitate is thus produced. Allow the suspension to stand in a boiling
water bath for 15 minutes, and then cool to room temperature. Prepare a paper-
pulp filter in a funnel fitted with a porcelain cone, and tamp well down. Decant
the clear solution through the filter and wash the precipitate twice by decantation
with about 20 cm^ of hydrochloric acid (1:9); this removes most of the excess of
quinoline and of molybdate. Transfer the precipitate to the pad with cold water,
washing the flask well; wash the filter and precipitate with 30-cm^ portions of
water, letting each washing run through before applying the next, until the
washings are acid free (test for acidity with pH test paper; about six washings are
usually required). Transfer the filter pad and precipitate back to the original flask:
insert the funnel into the flask and wash with about 50 cm^ of water to ensure the

315

X, 49 QUANTITATIVE INORGANIC ANALYSIS

transfer of all traces of precipitate. Shake the flask well so that filter paper and
precipitate are completely broken up. Run in 50.0 enf' of standard (carbonate-
free) 0.5Af-sodium hydro.xide. .swirling during the addition. Shake until the
precipitate is completely dissolved. Add a few drop.s of the mi.xcd indicator
solution and titrate with .standard 0.5A/-hydrochloric;icid to an end-point which
changes .sharply from pale green to pale yellow.

Rmi a blank on the reagents, but use O.LV-acid and alkali .solutions for the
titrations; calculate the blank to 0.5A/-sodium hydroxide. Subtract the blank
(which should not exceed 0.5 enr') from the volume neutralised by the original
precipitate.

1 cm’0,5A/-NaOH “ I.S30nic PO.,^“

Wilson (Ref. 12) has recommended that the hydrochloric acid added before
precipitation be replaced by citric acid, and the subsequent washing of the
precipitate is then carried out solely with distilled water.

The method can be applied to the determination of phosphorus in a wide
variety of materials, c.g.. phosphate rock, phosphatic fertilisers and metals, and is
suitable for u.se in conjunction with the oxygen flask procedure (Section III, 46).
In all eases it is essential to ensure that the material is .so trcjitcd that the
phosphorus is converted to orthophosphate; this may usually be done by
dissolution in tin oxidising medium such as concentrated nitric acid or in 60 per
cent perchloric acid.

B. 2 COMPLEXATION TITRATIONS

The simple complexation titration described in Section X, 22. i.c., the
determination of cyanide by titration with standard silver nitrate solution and
which involvc.s formation of the complex cyanoargcniatc ion [Ac(CN).]~. is most
conveniently included with the use of standard .silver nitrate solutions in
Precipitation Titrations (Section X. 87). Tlic follosving Sections will therefore be
devoted to applications of 1. 2-diaminoeilianctclra -acetic acid (cthylcnediamine-
tctra-acetic acid, EDF.A) and its congeners. Thc.sc reagents possess great
versatility arising from their inherent potency as complexing agents and from the
availability ofnumcrous nictal-ion indicator.s (Section X. 28). each effective over a
limited range of pll, but together covering a wide range of pll v.alucs: to these
factors must be added the additional refinements offered by 'masking' and
'demasking (echniques (Section X. 27).

It is clearly impossible, within the scope of the present volume, to give details
for all the caiion.s (and anions) which can be determined by EDTA or similar
types of titration. Accordingly, details of a few typical determinations arc given
which serve to illustrate the general proccdurc.s to be followed and the use of
various buflering agents and of .some different indicators. A conspectus of some
.selected proccdurc.s for the commoner cations is then given, followed by some
examples of the uses of EDTA for the determination of the components of
mixtures; finally, some examples of the determination of anions are given. The
relevant theoretical Sections (X, 19- 28) .should be consulted before commencing
the determinations.

316

TITRIMETRIC ANALYSIS X, 50/51

X, 50. STANDARD EDTA SOLUTIONS. Disodium dihydrogenethylene-
diaminetetra-acetate of analytical reagent quality is available commercially but
this may contain a trace of moisture. After drying the Analar material at 80 °C its
composition agrees exactly with the formula Na 2 H 2 CioHi 208 N 2 , 2 H 20
(molecular weight 372.24), but it should not be used as a primary standard. If
necessary, the commercial material may be purified by preparing a saturated
solution at room temperature: this requires about 20 g of the salt per 200 cm^ of
water. Add ethanol slowly until a permanent precipitate appears; filter. Dilute the
filtrate with an equal volume of ethanol, filter the resulting precipitate through a
sintered glass funnel, wash with acetone and then with diethyl ether. Air-dry at
room temperature overnight and then dry in an oven at 80 °C for at least 24
hours. Consult Chapter XXIII for a discussion of the thermal behaviour of
EDTA.

Solutions of EDTA of the following concentrations are suitable for most
experimental work; O.IM, 0.05M, and O.OIM and contain respectively 37.224 g,
18.612 g, and 3.7224 g of the dihydrate per dm^ of solution. As already indicated,
the dry Analar salt cannot be regarded as a primary standard and the solution
must be standardised; this can be done by titration of nearly neutralised zinc
chloride or zinc sulphate solution prepared from a known weight of A.R. zinc
pellets; nearly neutralised magnesium chloride (or sulphate) solution prepared
from a known weight of pure magnesium; or a manganese chloride solution
prepared from spectroscopically pure manganese.

The water employed in making up solutions, particularly dilute solutions, of
EDTA should contain no traces of polyvalent ions. The distilled water normally
used in the laboratory may require distillation in an all-Pyrex glass apparatus or,
better, passage through a column of cation exchange resin in the sodium form —
the latter procedure will remove all traces of heavy metals. De-ionised water is
also satisfactory; it should be prepared from distilled water since tap water
sometimes contains non-ionic impurities not removed by an ion exchange
column. The solution may be kept in Pyrex (or similar borosilicate glass) vessels,
which have been thoroughly steamed out before use. For prolonged storage in
borosilicate vessels, the latter should be boiled with a strongly alkaline, 2 per cent
EDTA solution for several hours and then repeatedly rinsed with de-ionised
water. Polythene bottles are the most satisfactory, and should always be
employed for the storage of very dilute (e.g., O.OOIM) solutions of EDTA. Vessels
of ordinary (soda) glass should not be used; in the course of time such soft glass
containers will yield appreciable amounts of cations (including calcium and,
magnesium) and anions to solutions of EDTA.

Water purified or prepared as described above should be used for the
preparation of all solutions required for EDTA or similar titrations.

X, 51. SOME PRACTICAL CONSIDERATIONS. The following points
should be borne in mind when carrying out complexometric titrations.

A. Adjustment of pH. For many EDTA titrations the pH of the solution is
extremely critical ; often limits of ± 1 unit of pH, and frequently limits of + 0.5 unit
of pH must be achieved for a successful titration to be carried out. To achieve
such narrow limits of control it is necessary to make use of a pH meter whilst
adjusting the pH value of the solution, and even for those cases where the latitude
IS such that a pH test paper can be used to control the adjustment of pH, only a
paper ofthe narrow range variety should be used. ■ .

317

TITRIMETRIC ANALYSIS X, 52/53

carried out stepwise, taking readings of the transmittance after each addition of
EDTA; these readings are then plotted against volume of EDTA solution added,
and at the end-point (where the indicator changes colour), there will be an abrupt
alteration in the transmittance, i.e., a break in the curve, from which the end-point
may be assessed accurately.

F. .Alternative methods of detecting the end-point. In addition to the visual
and spectrophotometric detection of end-points in EDTA titrations with the aid
of metal-ion indicators, the following methods are also available for end-point
detection!

1. Potentiometric titration using a mercury electrode (see Section XIV, 30).

2. Potentiometric titration using a selective ion electrode (Section XIV, 9-12)
responsive to the ion being titrated.

3. Potentiometric titration using a bright platinum-saturated calomel electrode
system; this can be used when the reaction involves two different oxidation
states of a given metal (see Section XIV, 31).

4. By conductometric titration (Chapter XV).

5. By amperometric titration (Chapter XVII).

6. By enthalpimetric titration (Chapter XXV).

A method of coulometric analysis (Chapter XIII) has also been devised; see
Reference 13.

Determination of Cations

X,51 DETERMINATION OF ALUMINIUM: BACK TITRATION
USING SOLOCHROME BLACK INDICATOR. Pipette 25 cm^ of an
aluminium ion solution (approximately 0.01 M) into a conical flask and run in
from a burette a slight excess of O.OIM-EDTA solution ; adjust the pH to between
7 and 8 by the addition of ammonia solution (test drops on phenol red paper or
use a pH meter). Boil the solution for a few minutes to ensure complete
complexation of the aluminium; cool to room temperature and adjust the pH to
7-8. Add 50 mg of Solochrome Black/KNOj mixture (see Section X, 51, C) and
titrate rapidly with standard O.OlM-zinc sulphate solution until the colour
changes from blue to wine red.

After standing for a few minutes the fully titrated solution acquires a reddish-
violet colour due to the transformation of the zinc dye complex into , the
aluminium-Solochrome Black complex; this change is irreversible, so that over-
titrated solutions are lost.

Every cm^ difference between the volume of O.OIM-EDTA added and the
O.OlM-zinc sulphate solution used in the back-titration corresponds to 0.2698 mg

The standard zinc sulphate solution required is best prepared by dissolving
a out 1.63 g (accurately weighed) of A.R. granulated zinc in dilute sulphuric acid,
nearly neutralising with sodium hydroxide solution, and then making up to 250
cm m a graduated flask; alternatively, the requisite quantity of A.R. zinc
su p ate may be used. In either case, de-ionised water must be used. ,

of BARIUM: DIRECT TITRATION WITH
^ INDICATOR. Pipette 25 cm^ barium ion

de-ir> A ^ 250-cm^ conical flask and dilute to about 100 cm^ with

nise water. Adjust the pH of the solution to 12 by the addition of 3-6 cm^ of

319

X, 54/55 QUANTfTATJVfi INORGANIC ANALYSIS

lAf sodium hydroxide solution; the pH must be checked with a pH meter as it
must lie between 11.5 and 12.7. Add 50 rng Methyl Thymol Bluc/KNOj (see
Section X, 51, C) mixture and titrate with standard (0.01 Af) EDTA solution until
the colour changes from blue to grey.

1 mole EDTA r= ImolcHa^*

X. 54. DETERMINATION OE BISMliTH: DIRECnTTRATION USING
xVlENOL ORANGE INDICATOR. Pipette 25 cm' of the bismuth solution
{approx. O.OI Af j into a 500 cm' conical flask and dilute with dc-ionised water to
.about 150 cm'. If necessary, adjust the pH lonbout I by the cautious addition of
dilute aqueous ammonia or of dilute nitric acid; use a pi I meter. Add .10 mg of the
Xylcnol Orangc/KNOj (see Section X, 51, C) mixture and then titrate with
standard 0.0JA/-EDTA solutioit until the red colour .starts to fade. From this
point add the tilrant slowly until the cnd-poinl is reached and the indicator
changes to yellow.

1 mole EDTA h 1 mole Hi' ‘

X, 55. DETERMINATION OF CALCIUM: SUHSTTTUTTON TITRA-
TION USING SOLOCHROME IH.ACK (ERIOCHROME BLACK T)
INDICATOR. DiM'ussion. When calcium ions arc titrated with EDTA a
relatively stable calcium comple.x is formed;

Ca'* h HjY*’- e±CaV'--2ir

With calcium ions alone, no sharp end-point can be obtained with Solochrome
Black (F.riochromc Bhrek T) indicator and the transition from red to pure blue is
not observed, With magnesium ions, a somewhat lc.ss stable complexonate is
formed :

Mg-’N ILY' jitMgY*’ +2ir

and the magnesium indicator complex is more .stable than the calcium-indicator
complex but less stable than the magnesium -EDTA complex. Consequently,
during the titration of a solution containing magnesium and c.alcium ions \siih
EDTA in the presence of Solochrome Black (Eriochromc Black T) the EDTA
reacts first with the free calcium ions, then with the free magnesium ions, and
finally with the magnesium -indicator complex. Since the magnesium-indicator
complex is wine red in colour and the free indicator is blue between pH 7 and 11.
the colour of the solution changes from wine red to blue at the end-point:

MgD - (red) -HI j Y' ' - M gY-’ ' f HD' - (blue) -t- 1 P

If magnesium ions arc not present in the solution containing calcium ions they
must be added, since they arc required for the colour change of the indicator. A
common procedure is (o add a small amount of magnesium chloride to the
EDTA solution before it is standardised. Another procedure, which permits the
EDTA solution to be used for other titrations, is to incorporate a little
rnagnesium-EDTA (MgY* “) (I 10 per cent) in the buffer solution or to add a
little 0.1 A/-magncsium-EDT.A (NajMgY) to the calcium-ion solution:

MgY'--fCa'" =CaY'-+Mg'’

Traces of many metals interfere in the determination of calcium and

320

TITRIMETRIC ANALYSIS X, 56

magnesium using Solochrome Black (Eriochrome Black T) indicator, e.g., Co, Ni,
Cu, Zn, Hg, and Mn. Their interference can be overcome by the addition of a little
hydroxylammoniura chloride (which reduces some of the metals to their lower
valency states) and also of sodium or potassium cyanide, which forms very stable
cyanide complexes (‘masking’). Iron may be rendered harmless by the addition of
a little sodium sulphide.

The titration with EDTA, using Solochrome Black (Eriochrome Black T) as
indicator, will yield the calcium content of the sample (if no magnesium is present)
or the total calcium and magnesium content if both metals are present. To
determine the individual elements, calcium may be evaluated by titration using a
suitable indicator, e.g., Patton and Reeder’s indicator or Calcon — see Sections X,
62 and X, 28, or by titration with EGTA using Zincon as indicator — see Section
X, 63. The difference between the two titrations is a measure of the magnesium
content.

Procedure. Prepare an ammonia-ammonium chloride buffer solution (pH
10), by adding 142 cm^ concentrated ammonia solution (sp. gr. 0.88-0.90) to
17.5 g A.R. ammonium chloride and diluting to 250 cm^ with de-ionised water.

Prepare the magnesium complex of EDTA, NajMgY, by mixing equal
volumes of 0.2M solutions of EDTA and of magnesium sulphate. Neutralise with
sodium hydroxide solution to a pH between 8 and 9 (phenolphthalein just
reddened). Take a portion of the solution, add a few drops of the buffer solution
(pH 10), and a few milligrams of the Solochrome Black (Eriochrome Black T)/
KNO 3 (see Section X, 51, C) indicator mixture. A violet colour should be
produced which turns blue on the addition of a drop of O.OIM-EDTA solution
and red on the addition of a single drop of 0 . 01 M-MgSO 4 solution ; this confirms
the equimolarity of magnesium and EDTA. If the solution does not pass this test,
it may be treated with more EDTA or with more magnesium sulphate solution
until the required condition of equimolarity is attained; this gives an
approximately O.IM solution. Alternatively solid Mg-EDTA complex may be
used, which is available from Hopkin and Williams Ltd.

Pipette 25.0 cin^ of the calcium-ion solution 0.01 M into a 250-cm^ conical
flask, dilute it with about 25 cm^ distilled water, add 2 cm^ buffer solution, 1 cm^
O.IM-Mg-EDTA, and 30-40 mg Solochrome Black (Eriochrome Black T)/
KNO 3 mixture. Titrate with the EDTA solution until the colour changes from
wine red to clear blue. No tinge of reddish hue should remain at the equivalence
point. Titrate slowly near the end-point.

1 mole EDTA = 1 mole Ca2 +

X, 56. DETERMINATION OF COPPER: DIRECT TITRATION USING
I'AST SULPHON BLACK F INDICATOR. This indicator is virtually
specific in its colour reaction with copper in ammoniacal solution; it forms,
CO oured (red) complexes with only copper and nickel, but the indicator action
with nickel is poor.

^'■ocerfure. Prepare the indicator solution by dissolving 0.5 g of the solid in
100 cm3 of de-ionised water.

d Pipette 25 cm^ of the copper solution (O.OIM) into a conical flask, add 100 cm^
j water, 5 cm^ concentrated ammonia solution and 5 drops of the

rtiJ or solution. Titrate with standard EDTA solution (O.OIM) until the colour

Changes from purple to dark green.

321

X. 57/58 QUANTITATIVE INORGANIC ANALYSIS

1 molcEDTA = I moIcCii-*

It should be noted that this method is only applicable to solutions containing
up to 25 mg copper ions in 1 00 cm-' of water; if the concentration of Cu^ ’ ions is
too iiigh, the intense blue colour of the coppcrfll) ammine complex masks the
colour change at the end-point. The indicsiior .solution mu.st be frc.shly prepared.

X, 57. DETERMINATION OF IRON(III): DIRECT TITRATION USING
VARIAMINE BLUE INDIC.VrOR. Procciliirc. Prepare the indicator solution
by dissolving ! g V'ariaininc Blue in 100 cm ' dc-ionised water; as alrcjidy pointed
out (Section X. 28). Variamine Blue acts a.s a redox indicator.

Pipette 25 cm' ironflll) solution t0.()5,\f) into a conical flask and dilute to 100
cm^ with dc-ioni,scd water. Adjust the pH to 2-3; Congo Red paper may be
u.scd-to the first perceptible colour change. Add 5 drops of the indicator
solution, warm the contents of the flask to 40 C. and titrate with standard
(0.05,\f) EDTA solution until the initial blue colour of the solution turns grcyjust
before the end-point, and with the final drop of reagent changes to yellow.

This particular titration is well adapted to be carried out potentiomctriaiilv
(Section XIV, 31).

1 mole EDTA == 1 mole Fc' '

X. 58. DETERMINATION OF NICKEL: DIRECT TITRATION.S USING
(a) MUREXIDE AND (h) BROMOPVROGALLOL RED AS INDl-
CATOR& Procedure (a). Prepare the indiaiior by grinding 0.1 g murexide
with 10 g ,A.R. potassium nitrate; u.se about 50 mg of the mixture for each
titration.

Also prepare a I Af sahttinn of ammonium chloride by dissolving 26.75 g of the
A.R. solid in de-ioni.sed water and making up to 500 cm' in a graduated flask.

Pipette 25 cm' nickel solution (0.01. \/) into a conical flask and dilute to 100
cm' with dc-ionised water. Add the solid indicator mixture (50 mg) and 10 cm' of
the l.\f ammonium chloride solution, and tiien .add concentrated ammonia
solution dropwisc until the pH is about 7 as shown by the yellow colour of the
solution. Titrate with standard (O.OIAf) EDTA solution until the end-point is
approached, then render the .solution strongly alkaline by the addition of 10 cm'
of concentrated ammonia .solution, and continue the titration until the colour
changes from yellow to violet. The pH of the final solution must be 10; at lower
pH values an orange-yellow colour develops and more ammonia solution must
be added until the colour is clear yellow. Nickel complexes rather slowly with
EDT.A. and consequently the EDTA .solution must be added dropwisc near the
end-point.

Procedure (h). Prepare the indicator by dissolving 0.05 g Bromopyrogallol
Red in 1 00 cm' of 50 per cent ethanol, and a bufftr solution by mixing 1(30 cm' of
lAf ammonium chloride solution with 100 cm' of lAf aqueous ammonia
solution.

Pipette 25 cm' nickel solution (O.OI.Af) into a conical flask and dilute to 150
cm with dc-ionised water. Add about 1 5 drops of the indicator solution. 10 cm'
of the bulTer solution and titrate with standard r£DTA solution (O.OIAf) un(il the
colour changes from blue to claret red.

1 mole EDT.A e 1 mole Ni' *

TITRIMETRIC ANALYSIS X, 59/60

X, 59. DETERMINATION OF SILVER: INDIRECT METHOD USING
POTASSIUM TETRACYANONICKELATE(n) AND NICKEL ION-
MUREXIDE INDICATOR.. Silver halides can be dissolved in a solution of
potassium tetracyanonickelate(II) in the presence of an ammonia-ammonium
chloride buffer, and the nickel ion set free may be titrated with standard EDTA
using murexide as indicator.

2Ag-^ +[Ni(CN)J^- :?^2[Ag(CN)J- +Ni"+

It can be shown from a consideration of the overall stability constants of the ions
[Ni(CN)4]^“(10^’) and [Ag(CN)2]“(10^^) that the equilibrium constant for the
above ionic reaction is 10^^, i.e., the reaction proceeds practically completely to
the right. An interesting exercise is the analysis of a solid silver halide, e.g.; silver
chloride. . • ■ ■ - , ■ . ■

Procedure. Prepare the murexide indicator as detailed in Section X, 58 [a),
and an ammonium chloride solution (IM) by dissolving 26.75 g A.R. ammonium
chloride in de-ionised water in a 500 cm^ graduated flask. , ■ , , ,

The potassium tetracyanonickelate(lI) which is required is prepared as follows.
Dissolve 25 g of A.R. NiS04,7H20 , in 50 cm^ distilled water and add
portion wise, with agitation, 25 g A.R. KCN. (Caution. Use a fume cupboard.) A
yellow solution forms and a white precipitate of K2SO4 separates. Gradually, add,
with stirring, 100 cm^ of 95, per cent ethanol, filter off the precipitated K2SO4
with suction, and wash twice with 2 cm^ ethanol. Concentrate the filtrate at about
70 °C — an infrared heater is convenient for this purpose. When crystals
commence to separate, stir frequently. When the crystalline mass becomes thick
(without evaporating completely to dryness), allow to cool and mix the crystals
with 50 cm^ ethanol. Separate the crystals by suction filtration and wash twice
with 5-cm^ portions ethanol. Spread the fine yellow crystals in thin layers upon
absorbent paper, and allow to stand for 2-3 days in the air, adequately protected
from dust. During this period the excess of KCN is converted into K2CO3'. The
preparation is then ready for use; it should be kept in a stoppered bottle.

Treat an aqueous suspension of about 0.072 g (accurately weighed) silver
chloride with a mixture of 10 cm^ of concentrated ammonia solution and 10 cm^
of IM ammonium chloride solution, then add about 0.2 g of potassium
cyanonickelate and warm gently. Dilute to 100 cm^ with de-ionised, water, add 50
mg of the indicator mixture and titrate with standard (O.OIM) EDTA solution,
adding the reagent dropwise in the neighbourhood of the end-point, until the
colour changes from yellow to violet.

1 mole EDTA = 2 moles Ag'*’

Palladium(II) compounds can be determined by a similar procedure, but in this
case, after addition of the cyanonickelate, excess of standard (O.OIM) EDTA
solution is added, and the excess is back titrated with standard (O.OIM)
manganese(II) sulphate solution using Solochrome Black indicator.

Gold may be titrated similarly.

X, 60. DETERMINATION OF SODIUM: INDIRECT TITRATION
USING SOLOCHROME BLACK (ERIOCHROME BLACK T) INDI-
sodium is precipitated as sodium zinc uranyl acetate NaZn
1 y 02)3(CH3C00)9,6H20 (Section XI, 53, B), and the zinc is then determined bv
titration with EDTA.

323

X, 61 QUANTITATIVE INORGANIC ANALYSIS

Procedure. Prepare the indicator by grinding 0.1 g of Solochrome Black
(Eriochrome Black T) with 10 g A.R. potiLSsiiim nitrate, a buffer .'solution (pH 10)
as in Section X. 55, an ammonium carbonate solution ( 1 M) by dissolving 78.6 gof
the A.R. solid in 500 effl^ of dc-ionised water, and the predpiiatum reagent as
detailed in Section XL 53. B: hydrochloric acid {! A/) will also be required.

The solution which contains not more than 5 mg of sodium must be
concentrated to a small bulk (1-2 cm"), and precipitation carried out as detailed
in Section XI. 53, B. After filtering through a porcelain filter crucible the
precipitate of zinc uranyl acetate is washed four limes with 2 cm-* portions of the
precipitating reagent, and finally with ten 2 cm^ portions of 95 per cent ethanol
which has licen saturated with sodium zinc uranyl acetate. The filter crucible is
then stood inside a -KK) cm’ Pyrex beaker and 5 cm ’ of 1 Af-hydrochloric acid
added to the crucible. After a few minutes add 50 cm’ of dc-ionised water and
boil. Allow the solution to cool somewhat, remove the crucible with crucible
tongs and wash carefully into the beaker. Then set up the crucible again for
filtration and wash the sinter thoroughly. Neutralise the combined filtrate and
washings (total volumeabout lOOcm’lwith !\f-ammoniumcarbonaleand.ndd2
cm’ in excess to hold the uranium in solution as the carbonato-complex. Add 2
cm-' liuITer mixture (pH - 10) and .30 mg of Solochrome Black (Eriochrome
Black T/KNOj) indicator. Titrate with standard 0.001 Af-EDTA. 'ntc colour
change at the end-point is from yellovsish-rcd to grcenish-blue.

1 mole HDTA s: 1 mole Na *

Sodium, free from all other cations, may be determined by passage through a
cation exchange column in the magnesium form and titration of the liberated
magnesium ion with .standard EDTA solution.

X, 61. DCTAIIS FOR TME DO ER.Mf. NATION OF A SELECTION OF
METAL IONS BY EDTA TITR.ATION. With the dcl;iilcd instructions given
in Sections X. 52-60 it should be possible to carry out any of the following
determinations in Table X. 13 without serious problems arising. In all cases it is
recommended that the requisite pH value for the titration should be established
by use of a pH meter, but in the light of experience the colour of the indicator at
the required pH may. in some case,s, be a satisfactory guide. Where no actual
buffering agent is specified, the solution should be brought to the required pH
value by the cautious addition of dilute acid or of dilute sodium hydroxide
solution or aqueous ammonia solution as required.

Table X, 13. Summarised procinlures for EDTA titrations of some .selected
cations

.Mclal

Titration

IlufTer

Indicator

Colour

Notes

l)-pc

(I)

chanjtc

(2)

Aluminium*

D.icr

7-S

Aq.NHj

fiarium*

Direct

MTB

(3)

Bismuth

Direct*

Direct

O-I

MTB

Cadmium

Direct

Hcxaminc

Calcium

Direct

MTB

Substn.*

7-11

Aq NHj/NH^CT

,SB

Cob.ill

Direct

llcjtaminc

f'J

TITRIMETRIC ANALYSIS X, 62

Table X, 13.

Metal

Titration

type

Buffer

Indicator

(1)

Colour

change

(2)

Notes

Copper*

Direct

FSB

(6)

Gold

See silver

Ironail)* ■

Direct

2-3

( 4 )

Lead

Direct

Hexamine

Magnesium

Direct

Aq.NH3/NH4Cl

SB ,

( 4 *) .

Manganese

Direct

Aq.NH3/NH4Cl

( 5 )

Direct

■ 10

Aq.NH3

TPX

( 5 )

Mercury

Direct

Hexamine

Direct

Hexamine

MTB

Nickel

Direct*

7-10

Aq. NH3/NH4CI

, M

Direct*

7-10

Aq. NH3/NH4Ci

BPR

Back

Aq.NH3/NH4Cl

Palladium

See silver

Silver*

Indirect

Sodium*

Indirect

Aq. NH3/NH4CI

Strontium

Direct

MTB

Direct

10-11

TPX

Thorium

Direct

2-3

Direct

2-3

MTB

Direct

3 - 3.5

( 4 )

Tin(II)

Direct

Hexamine

Zinc

Direct

Aq.NH3/NH4CI

Direct

Hexamine

Direct

Hexamine

MTB

Notes to Table X, 15.

‘/Details in Sections X, 52 - 60 .

( 1 ) BPR = Bromopyrogallol Red; CV = Catechol Violet; FSB = Fast Sulphon Black F;
M = Murexide; MTB = Methylthymol Blue; SB = Solochrome Black (Eriochrome Black T);
TPX = Thymolphthalexone; VB = Variaraine Blue; XO = Xylenol Orange.

( 2 ) B = Blue; G = Green; Gr. = Grey; O = Orange; P = Purple; PP = Pale pink; R = Red; V
= Violet; Y = Yellow.

( 3 ) Can also be determined by precipitation as BaS04 and dissolution in excess EDTA (Section X,
75 ).

( 4 ) Temperature 40 C; ( 4 *) Warming optional.

( 5 ) Add 0.5 g hydroxylammonium chloride (to prevent oxidation), and 3 cm^ triethanolamine (to
prevent precipitation in alkaline solution); use boiled-out (air-free) water.

(6) In presence of concentrated aqueous ammonia.

Analysis of mixtures of cations

X, 62, DETERMINATION OF CALCIUM AND MAGNESIUM USING
PATTON AND REEDER’S INDICATOR. Discussion. Patton and
Reeder’s indicator (HHSNNA), see Section X, 28, permits the determination of
calcium in the presence of magnesium, and finds application in the determination
of the hardness of water and in the analysis of limestone and dolomite. Titration
using Solochrome Black (Eriochrome Black T) gives calcium and magnesium
together, and the difference between the two titrations gives the magnesium
content of the mixture (1).

Calcon may also be used for the titration of Ca in the presence of Mg (compare
Section X, 28). The neutral solution (say, 50 cm^) is treated with 5 cm^ of
diethylamine (giving a pH of about 12.5, which is sufficiently high to precipitate

325

X, 62 QUANTITATIVE INORGANIC ANALYSIS

the magnesium quantitatively as the hydroxide) and 4 drops of Calcon indicator
arc added. The .solution is stirred magnetically and titrated with standard EDTA
solution until the colour changes from pink to a pure blue.

A sharper end-point may l>c obtained by adding 2-3 drops of I per cent
aqueous polyvinyl alcohol to the sample solution, then adjusting the pH to 12.5
with sodium hydroxide, adding 2-3 drops of 10 per cent aqueous potassium
cyanide solution, warming to 60 ’ C (Caulion: use a fume cupboard), and treating
the warm solution with 3-4 drops of Calcon indicator. The solution is titrated
with O.OIM-EDTA to a red-blue end-point. The polyvinyl alcohol reduces the
adsorption of the dye on the surfacx- of the precipitate. 'Tlic solution is prepared by
mixing 1.0 g of med'ium-vi.sco.siiy polyvinyl alcohol with 100 em^ of boiling water
in a mechanical homogeni/cr.

Procedure. Prepare the ir.dictUor.K by grinding (n) 0.5 g HHSNNA with 50g
A.R. potassium chloride, and (h) 0.2 g Solochrome Black (Eriochromc Black T)
with 50 g .A.R. potassium chloride. Hie following solutions will also be required;

A/ug/ie.vhw! chloride .solution (0.01. \I). Dissolve 0.60S g pure magnesium
turnings in dilute hydrochloric acid, nearly neutralise wdih sodium hydroxide
solution (lA/) and make up to 2.^0 cni^ in a graduated flask with de-ionised water.
Pipette 25 cm’ of the rc.sulting 0. 1 M solution into a 250 cm’ graduated flask and
make up to the mark with de-ioni<-cd water.

Poto.ssiiuu hydroxide .solution ca. 8.M. ETissolvc 112 g A.R. potassium
hydroxide [wilcts in 250 cm’ of de-ionised water.

Buffer .solution. Add 55 cm ' concentrated hyiirochloric acid to 400 cm’ de-
ionised water vind mix thoroughh. Slowly pour 310 cm’ redistilled mono-
cthanolaminc with stirring into the mixture and cool to room temperature (2).
Titrate 50.0 cm’ of the standard ,MgCK solution with standard (0.01 .\/) P.DT.A
solution trsing 1 cm 'oflhemonocthanolamine hydrochloric acid solution as the
bufl'er .and Solochrome Black (l;riochromc Black T) as the indictitor. .Add 50.0
cm’ of the NtgCl; solution to the volume of F.DTA solution required to complex
the magnc.sium exactly (as determined m the last titration), pour the mixture into
the monoethanolaminc hvdrochloric acid solution, and mix \s ell. Dilute to 1 litre
(3).

Determinatiou of euleium. Pipette two 25.0 cm’ portions of the mixed
calcium- and niagncsium-ion solution (not more than 0.01. \f witli respect to
either ion) into two scpiiratc 250-cm’ conical flasks and dilute each with about 25
cm’ of distilled water. To the first flask add 4 ern’ 8M-polassium hydroxide
solution {a precipitate of maciicsium hydroxide may be noted here), and allow to
stand for 3-5 minutes with occasion.ii .swirling. Add about 30 mg each of A.R.
potassium cyanide (Caution: poison i .ind A.R. hydroxylaminonium chloride and
swirl the contents of (he flask uniil the soiid.s dissolve. Add about 50 mg of the
HHSNNA indicator mixtur. and titrate with 0.0lAf-l:DTA until the colour
changes from red to blue. Run into the .second flask from a burette a volume of
EDTA solution equal to tliat required to reach theend-point lc.s.s 1 cm’. Nowadd
4 cm* of the potassium hyiiroxidc solution, mix well and complete the titration as
with the fir.st sample; record the exact volume of EDTA solution u.scd. Perform a
blank titration, replacing the sample with distilled water.

Detenniuotiou of total calcium and moy,ne,siunt. Pipette 25.0 cm’ of the
mixed calcium- and magnesium-ion solution into a 250-cm’ conical flask, dilute
to about 50 cm’ with distilled water, add 5 cm’ of the buffer solution, and mix by
swirling. Add about 30 mg A.R. potassium cyanide and A.R. hydroxylammonium

TITRIMETRIC ANALYSIS X, 63

chloride, shake gently until the solids dissolve, and then add about 50 mg of the
Solochrome Black (Eriochrome Black T) indicator mixture. Titrate with the
EDTA solution to a pure blue end-point. Perform a blank titration, replacing the
25-cm^ sample solution with de-ionised water.

Calculate the volume of standard EDTA solution equivalent to the magnesium
by subtracting the, total volume required for the calcium from the volume
required for the total calcium and magnesium for equal amounts of the test
sample.

Notes. 1. The usefulness of the HHSNNA indicator for the titration of
calcium depends upon the fact that the pH of the solution is sufficiently high to
ensure the quantitative precipitation of the magnesium as magnesium hydroxide
and that calcium forms a more stable complex with EDTA than does magnesium.
The EDTA does not react with magnesium (present as Mg(OH) 2 ) until all the free
calcium and the calcium-indicator complex have been complexed by the EDTA.
If the indicator is added before the potassium hydroxide, a satisfactory end-point
is not obtained because magnesium salts form a lake with the indicator as the pH
increases and the magnesium indicator-lake is coprecipitated with - the
magnesium hydroxide.

2. The monoethanolamine-hydrochloric acid buffer has a buffering capacity
equal to the ammonia-ammonium chloride buffer commonly employed for the
titration of calcium and magnesium with EDTA and Solochrome Black
(Eriochrome Black T) (compare Section X, 55). The buffer has excellent keeping
qualities, sharp end-points are obtainable, and the strong ammonia solution is
completely eliminated.

3. When relatively pure samples' of calcium are titrated using Solochrome
Black (Eriochrome Black T) as indicator, magnesium must be added to obtain a
sharp end-point, hence magnesium is usually added to the buffer solution
(compare Section X, 55). The addition of magnesium to the EDTA solution
prevents a sharp end-point when calcium is titrated using HHSNNA as indicator.
The introduction of complexed magnesium into the buffer eliminates the need for
two EDTA solutions and ensures an adequate amount of magnesium, even when
small amounts of this element are titrated.

X, 63. DETERMINATION OF CALCIUM IN THE PRESENCE OF
magnesium using EGTA as TITRANT. Discussion. Calcium niay be
determined in the presence of magnesium by using EGTA. as titrant, because
whereas the stability constant for the calcium-EGTA complex is about 1 x 10^ *,
that of the magnesium-EGTA complex is only about 1x10^, and thus
magnesium does not interfere with the reagent. The method described in' the
preceding section, which involves precipitation of magnesium hydroxide, is not
satisfactory if the magnesium content of the mixture is much greater than about
10 per cent of the calcium content, since co-precipitation of calcium hydroxide
may occur. Titration with EGTA is therefore to be recommended for the
determination of small amounts of calcium in the presence of larger, amounts of
magnesium.

The indicator used in the titration is Zincon (Section X, 28) which gives, rise to
an indirect end-point with calcium. Detection of the end-point is dependent upon
the reaction

ZnEGTA2-+Ca2+ =Zn2+-fCaEGTA2- ,

327

X, M QUANTITATIVE INORGANIC ANALYSIS

and the zinc ions liberated form a blue complex with the indicator. At the end-
point. the zinc-indicator complex is decomposed;

Znln-~ + I-GTA - ZnEGTA’" + Min'*"

and the solution acquires the orange-red colour of the indicator.

Promlwr. Prepare an EGr/l .so/i/nVwifO.OSAf) by dissolving 19.01 gin 100
cm^ sodium hydro.xidc solution ( 1 A/) and diluting to 1 dm-’ in a graduated ilask
with de-ionised water. Prepare the iiulinilar by dissolving 0.005 g Zincon in 2 era-’
sodium hydroxide solution ( 0. 1 ^^) and diluting to 1 00 cm-’ with de-ionised water,
and a buffer snliilum {pH 10) by dissolving 25 g sodium tetraborate, 3.5 g
ammonium chloride, and 5.7 g sodium liydroxide in ! litre of de-ionised water.

Prepare 100 cm-’ of Zv-KGTA complex solution by taking 50 cm’ of O.OSAf
zinc sulphate solution and adding an equivalent volume of O.OSAf EGTA
solution: exact equality of zinc and EGTA is best achieved by titriiting a 10 cm^
portion of the zinc sulphate solution with the EGT.A solution using zincon
indicator, and from this result the exact volume of EGTA solution required for
the 50 cm’ portion of zinc .sulphate solution may be calculated.

The FAiTA solution muy he staiuhirtlisiul by titnilion of a standard (O.OSAf)
calcaum .solution, prepared by dissolving 5.00 g A.R. calcium carbonate in dilute
hydrochloric acid contained in a dm-’ gnidiiatcd Ilask. and then after neutralising
w-ith sodium hydroxide solution diluting to the m:irk with de-ionised water; use
zincon indicator in the presence of Zn EGT.A solution (sec below-).

T o deicnninc the rolcium in the calcium -magne.sium mixture, pipette 25 cm’ of
the solution into a 250-cm-' conicid flask, add 25 cm’ of the bulTer .solution and
check that the re.sulting ,':o!u(ion has a pll of 9.5 -10, 0. Add 2 cm-’ of the Zn-
EGTA .solution and 2-3 drops of (he indicator solution. Titrate .slow-)y with the
standard EGTA solution until the blue colour changes to orange-red.

X,64. DETEKMI.N'.ATlONOFTfli: TOTAL HARDM-SS(PERMANENT
AND TEMPORARY) OF \V.\TER USING SOEOCHROME BLACK
(ERIOCTIRO.ME BI.ACK I) INDICATOR. The hardness of water is
generally due to dissolveti calcium and magnesium salts and may be determined
by complexometne titration.

Procedure. To a 50 cm -' s;implc of the water to Ik' tested add 1 cm’ buffer
solution (aq.NI 1 3 ,'NH jCl, pH 10 , Section X, 55) ;ind 30 - 40 mg Solodirome Black
(Eriochromc Black 'lyKNOj indicator mixture. Titrate with standard EDTA
solution (0.01 Af) until the colour changes from red to pure blue. .Should there be
no magnesium prc.scnt in the sample of water it is necessary to add 0,1 cm’
magnesium -EDTA solution (0.1 Af ) before adding the indicator (see Section X,
55). The total hardness is expre.sscd in parts of CaCO , per million of w ater.

If the water contains traces of interfering ions, then 4 cm-’ of bulTcr solution
should be added, followed by 30 mg hydro.xylammoniiim chloride and then 50
mg A.R. potassium cyanide (Caution) Inrforc adding the indicator.

Notes. 1 . Somew-hat sharper end-points may be obtained if the sample of
Witter is first acidified with dilute hydrochloric acid, boiled for about a minute to
drive off carbon dioxide, cooled, neutralised with sodium hydroxide solution,
buffer and indicator solution added, and then titrated with EDTA as above.

2. The permanent hardne.ss of a sample of w-atcr may be determined as follows.
Place 250 cm’ of the sample of water in a 600-cm’ beaker and boil gently for 20-
30 minutes. Cool and filter it directly into a 250-cm’ siradu-atcd flask ; do not w-ash

328

TITRIMETRIC ANALYSIS X, 65/66

the filter paper, but dilute the filtrate to volume with distilled water and mix well.
Titrate 50.0 cm^.of the filtrate by the same procedure, as was used for the total
hardness. This titration measures the permanent hardness of the water. Calculate
this hardness as parts per million of CaCOj. ■

Calculate the temporary hardness of the water by subtracting the permanent
hardness from the total hardness. . . ' • ;

3. If it is desired to determine both the calcium and the magnesium in a sample
of water, determine first the total calcium and magnesium content as above, and
calculate the result as parts per million of CaCOj.

The calcium content may then be determined by titration with EDTA using
either Patton and Reeder’s indicator or Calcon {Section X, 62), or alternatively by
titration with EGTA (see previous Section). •

X, 65. DETERMINATION OF CALCIUM IN THE PRESENCE OF
BARIUM USING CDTA AS TITRANT. There is an appreciable difference
between the stability constants of the CDTA complexes of barium (logK = 1.99)
and calcium (logK = 12.50), with the result that calcium may be titrated with
CDTA in the presence of barium; the stability constants of the EDTA complexes
of these two metals are too close together to permit independent titration of
calcium in the presence of barium.

The indicator Calcichrome (see Section X, 28) is specific for calcium at a pH 1 1-
12 in the presence of barium.

Procedure. Prepare the CDTA solution (0.02M) by dissolving 6;880 g of the
solid reagent in 50 cm^ of sodium hydroxide solution (IM) and making up to 1
dm^ with de-ionised water; the solution may be standardised against a standard
calcium solution prepared from 2.00 g A.R. calcium carbonate (see Section X, 63).
The indicator is prepared by dissolving 0.5 g of the solid in 100 cm^. of water.

Pipette 25 cm^ of the solution to be analysed into a 250 cm^ conical flask and
dilute to 100 cm^ with de-ionised water: the original solution should be about
0.02M with respect to calcium and may contain barium to a concentration of up
to 0.2M. Add 10 cm^ sodium hydroxide solution (IM) and check that the pH of
the solution lies between 11-12; then add 3 drops of the indicator solution.
Titrate slowly with the standard CDTA solution until the pink colour changes to
blue.

X, 66. DETERMINATION OF CALCIUM AND LEAD IN ADMIX-
TURE USING METHYLTHYMOL BLUE INDICATOR. With Methyl-
thymol Blue lead may be titrated at a pH of 6 without interference by calcium; the
calcium is subsequently titrated at pH 12.

Procedure. Pipette 25 cm^ of the test solution (which may contain both
calcium and lead at concentrations of up to O.OIM) into a 250 cm^ conical flask
and dilute to 100 cm^ with de-ionised water. Add about 50 mg of the
MTB/KNO3 mixture followed by dilute nitric acid until the solution is yellow,
and then add powdered hexamine until the solution has an intense blue colour
(pH ca. 6). Titrate with standard (O.OIM) EDTA solution until the colour turns to
yellow; this gives the titration value for lead.

Now carefully add sodium hydroxide solution (IM) until the pH of the solution
has risen to 12 (pH meter); 3-6 cm^ of the sodium hydroxide solution will be
required. Continue the . titration of the bright blue solution with the EDTA
solution until the colour changes to grey ; this ^he titration value for calcium.

X, 67 QUANTITATIVB INORGANIC ANALYSIS

X 67 determination OF MAGNESIUM, MANGANESE, AND
ZINC IN ADMIXTURE: USE OF FLUORIDE ION AS A DEMASKING
AGENT. Discussion. In mixtures of m.timesium and manganese the sum of both
ion concentrations may be determined by direct EDTA titration. Fluoride ion
will demnsk magnesium selectively from it.s EI3TA complex, and if excess of a
standard solution of manganese ion is also added, the following reaction occurs
at room temperature:

MgY-" +2F" -t-Mn-’ “ Mgr,-*-MnY-'

The excess of manganese ion is csahiaied by back -titration with EDTA. The
amount of standard manganese ion solution consumed is equivalent to the
EDT.A ‘liberated’ by the fluoride ion. which i.s in turn equivalent to the
magnesium in the sample.

Mi.xtures of mane.tncse. magnesium, and 7inc cxin be similarly anaiy,scd. The
first EDTA end-point gives the sum of the three ions. Fluoride ion is added and
the EDT.A liivcraled from (he magnesium -EDT.A complex is titrated with
manganese ion as detailed above. Following the second end-point, cyanide ion is
added to displace zitic from its EI3TA chelate and to form the stable cyanor.incate
complex |Zn(CN)j]‘ " ; the liberated EDT A (equivalent to the y.inc) is titrated with
standard manganese-ion solution.

Details for the anulvMs of Mn -Mg-Zn mixtures will be given.

PriKCiiurc. Prepare a imincuncMill) yjlpluuc soluiirm (approx. 0.05, \/) by-
dissolving 11.15 g of the .A.R. solid in I litre of dc-ionised water; standardise the
solution by titration with 0,05.5/-EDTA solution using .Solochrome Black
indicator after the addition of 0.25 g hydroxylammonium chloride- see below.

Prepare a Iniljcr .solution \p!l 10) by drssoh ing 8,0 g ,A.R. ammonium nitrate in
65cnr' ofde-ionised w-.ater and adding .55 cm' of concent rated ammonia solution
(sp. gr. 0.88).

Pipette 25 enr' of the solution containing magnesium, manganese and zinc ions
(each appro,x. 0,02, \f). into a 250 can’ conical flask and dilute to 100 enr' with de-
ionised water. Add 0.25 g hydroxylammonium chloride (this is to prevent
o.xidation of Mn(H)ion.s). followed by 10 cm’ of the buffer solution and 30-40 mg
of the indieator/'KNOj mi.xturc. Warm to 40 C and titrate (preferably stirring
magnetically) with the standard EDTA solution to a pure blue colour.

.After the end-point, add 2,5 g of sodium fluoride and stir for agitate) for 1
minute. Now introduce She standard manganese') 11) sulphate solution from a
burette in I-cm’ portions until a penr, incut red colour is obtained; note the exact
volume added. Stir for 1 minute. Titrate the c.xccss of manganese ion with EDTA
until the colour changes to pure blue.

After the second end-point, add 4-5 cm’ of 15 per cent aqueous potassium
cyanide .solution, and run in the standard manganese-ion solution from a burette
until the colour changes sharply from blue to red. Record the exact volume of
manganc.scfll) sulphale, solution added.

Calculate the weights of magnesium, zinc, and manganese in the sample
solution.

, A.v<wip/e o/ calculation. In the standardisation of tlie Mnfll). solution, 25.0
cm- of the solution required 30,30 cm’ of 0,0459.\f-EDTA solution.

Molarity of Mn(H), solution = (30,30 x O.IMSOX'^S.O = 0.0556,5/
hirst titration of mixture with EDTA{Mg-(-Mn-hZn) = 33.05 cm
Second titration (afteraddincNaF; gives Mg) 9.S5cnr’ofMn(lI).solution

330

TITRIMETRIC ANALYSIS X, 68/69

and the excess Mn(II) required 1.26 cm^ of standard EDTA solution.

■ Millimoles ofEDTA liberated by NaF

= (9.85 X 0.0556) -(1.26 X 0.0459)
= 0.4899

and weight of magnesium per cm^ = (0.4899 x 24.3 l)/1000g

= 11.91 mg cm“^

Third titration (after adding KCN ; gives Zn) = 8,46 cm^ of Mn(II) solution.

■ Millimoles ofEDTA liberated by KCN

= (8.46 X 0.0556)

= 0.4703

and weight of zinc per cm^ = (0.4703 x 65.38)/1000 g

= 30.75 mgcm“^

In the first titration, (33.05 x 0.0459) millimoles of EDTA were used

= 1.5170 millimoles,

which represents the total amount of metal ion titrated (Mn + Mg + Zn).

Hence amount of Mn = 1.5170— (0.4899 + 0.4703) = 0.5568 millimoles

and weight of manganese per cra^ = (0.5568 x 54.94)/1000 g

= 30.60 mg cm

X,68. DETERMOINATION OF CHROMIUM(III) AND ffiON(m) IN
ADMIXTURE; AN EXAMPLE O F KINETIC MASKIN G. Iron (and nickel,
if present) can be determined by adding an excess of standard EDTA to the cold
solution, and then back-titrating the solution with lead nitrate solution using
Xylenol Orange as indicator; provided the solution is kept cold, chromium does
not react. The solution from the back titration is then acidified, excess of standard
EDTA solution added and the solution boiled for 15 minutes when the red-violet
Cr(III)-EDTA complex is produced. After cooling and buffering to pH 6, the
excess EDTA is then titrated with the lead nitrate solution.

Procedure. Place 10 cm^ of the solution containing the two metals (the
concentration of neither of which should exceed 0.01 M) in a 600 cm^ beaker fitted
with a magnetic stirrer, and dilute to 100 cm^ with de-ionised water. Add 20 cm^
of standard (approx. 0.01 M) EDTA solution and add hexamine to adjust the pH
to 5-6. Then add a few drops of the indicator solution (0.5 g Xylenol Orange
dissolved in 100 cm^ of water) and titrate the excess EDTA with a standard lead
nitrate solution (0.01 Af), i.e., to the formation of a red-violet colour.

To the resulting solution now add a further 20 cm^ portion of the standard
EDTA solution, add nitric acid (IM) to adjust the pH to 1-2, and then boil the
solution for 15 minutes. Cool, dilute to 400 cm^ by the addition of de-ionised
water, add hexamine to bring the pH to 5-6, add more of the indicator solution,
and titrate the excess EDTA with the standard lead nitrate solution.

The first titration determines the amount ofEDTA used by the iron, and the
second, the amount of EDTA used by the chromium.

X, 69. DETERMINATION OF MANGANESE IN PRESENCE OF IRON-
ANALYSIS OF FERRO-MANGANESE. After dissolution of the alloy in a
mixture of concentrated nitric and hydrochloric acids the iron is masked with
triethanolamine in an alkaline medium, and the manganese titrated with
standard EDTA solution using Thymolphthalexone as indicator. The amount of
iron(III) present must not exceed 25 mg per 100 cm^ of solution otherwise the
colour of the iron(III)-triethanolamine complex is rn. intense that the colour

331

X, 70 QUANTITATIVE INORGANIC ANALYSIS

change of the indicator is obscured. Consequently, the procedure can only be
used for samples of ferro-mancanesc containing more than about 40 per cent
mnneanese.

Procedure. Dissolve a weighed amount of fcrro-manganc.se {about 0,40 gl
in concentrated nitric acid and then add concentrated hydrochloric add (or use a
mixture of the two concentrated acids); prolonged boiling may be necessary.
Evaporate to a small volume on a water bath. Dilute with water and filter directly
into a lOO-cnr’ graduated hask. wash with distilled water and finally dilute to the
mark. I'ipcttc 25,0 cm^ of the solution into a .ISO-cnr' conical flask, add 5 ern’ of
10 per cent aqueous hydro.xylammoniuni chloride solution, 10 enr’ of 20 per cent
aqueous triethanolamine solution. 30 35 cm-' of concentrated ammonia
solution, tibout 100 cm' of water, and 6 drops of Thymolphthalexone indicator
solution. Titrate with standard O.OS.M-EDTA until the colour changes from blue
to colourIcs,s (or a very pale pink).

X,70. DETER Ml. NATION OE NICKEL IN PRESENCE OF IRON:
ANALYSIS OF NICKEL STEEI_ Nickel may be determined in the presence
of a large excess of irondlll in weakly acidic solution by adding EDTA and
triethanolamine; the intense brown precipitate di.ssolves upon the addition of
aqueous sodium hydroxide to yield a colourless solution. The iron(lll) is present
as the triethanolamine complex and only the nickel is complexed by the EDTA.
The excess of liDTA is back-titrated with .standard calcium chloride solution in
the presence of Thy molpluhalexonc indicator. The colour change is from colour-
less or very pale blue to ati intense blue. The nickel -EDTA complex has a faint
blue colour; the solution should contain le.ss than .35 mg of nickel per 100 cm'.

In the back-titration small amounts of copper and zinc and trace amounts of
manganese are quantitatively displaced from the EDTA and arc complexed by
the triethanolamine; smalt quantities of cobalt are converted into a
triethanolamine complex during the titration. Relatively high concentrations of
copper can be masked in the alkaline medium by the addition ofthioglycollicacid
until colourless, .Manganese, if present in quantities of more than 1 mg may be
oxidised by air and form.s a mangancscdll) irictli.'inolamine comple.x, which is
intensely green in colour; tins docs not occur if a little hydroxylammonium
chloride solution is added.

Procedure. Prepare a sitmdard coleiuoi chloride solution {O.Ol.M) by
dissolving 1.000 g of A.R. calcium v-arbonaic in the minimum volume of dilute
hydrochloric acid and diluting lo I dnv' with dc-ionised water in a graduated
flask. Also prepare a 20 per cent aqueous solution of triethanoUtmine.

Weigh out accurately a 1.0 g sample of the nickel steel and dissolve it in the
minimum volume of concentrated hydrochloric acid (about 15 cur') to whicii a
little concentrated nitric acid (co. I cm') has l>een added. Dilute to 250 cm' in a
graduated (lask. Pipette 25,0 cm'' of this- soluiion into a conical flask, add 25.0 cm'
of 0.0Ly/-EDTA and 10 enr' of triethanolamine solution. Introduce lAf-sodium
hydroxide solution, with stirring, until the pH of the solution is 1 J.6 (use a pH
meter). Dilute to about 250 cm'. Add about 0.05 g of the indicator-KNOj
mixture; the solution acquires a very pale blue colour. Titrate with O.Ol.Af-
calcium chloride solution until the colour changes to an intense blue. If it is felt
that the end-point colour change is not siidicicntlv distinct, add a further small
amount of the indicator, a known volume of 0.01 ANEDTA and titrate again with
0.01 A/-calcium chloride.

332

TITRIMETRIC ANALYSIS X, 71/72

X,71. DETERMINATION OF LEAD AND TIN IN ADMIXTURE:
ANALYSIS OF SOLDER. A mixture of tm(IV) and lead(II) ions may be
complexed by adding an excess of standard EDTA solution, the excess EDTA
being determined by titration with a standard solution of lead nitrate; the
total lead plus tin content of the solution is thus determined. Sodium fluoride is
now added and this displaces the' EDTA from the tin(I'V)-EDTA complex; the
liberated EDTA is determined by titration with a standard lead solution.

Procedure. Prepare a standard EDTA solution (p.2M), a standard lead
solution {O.OIM) ', a 30 per cent aqueous solution of hexamine, and a 0.2 per cent
aqueous solution oiXylenol Orange.

Dissolve a weighed amount (about 0.4 g) of solder in 10 cm^ concentrated
hydrochloric acid and 2 cm^ concentrated nitric acid; gentle warming is
necessary. Boil the solution gently for about 5 minutes to expel nitrous fumes and
chlorine, and allow to cool slightly, whereupon some lead chloride may separate.
Add 25.0 cm^ of standard 0.2M-EDTA and boil for 1 minute; the lead chloride
dissolves and a clear solution is obtained. Dilute with 100 cm^ of de-ionised
water, cool and dilute to 250 cm^ in a graduated flask. 'Without delay, pipette two
or three 25.0 cm^ portions into separate conical flasks. To each flask add 15 cm^
hexamine solution, 110 cm^ de-ionised water, and a few drops of Xylenol Orange
indicator. Titrate with the standard lead nitrate solution until the colour changes
from yellow to red. Now add 2.0 g A.R. sodium fluoride; the solution acquires a
yellow colour owing to the liberation of EDTA from its tin complex. Titrate again
with the standard lead nitrate solution until a permanent (i.e., stable for 1 minute)
red colour is obtained. Add the titrant dropwise near the end-point; a temporary
pink or red colour gradually revertirfg to yellow signals the approach of the end-
point.

X, 72. DETERMINATION OF BISMUTH, CADMIUM AND LEAD IN
ADMIXTURE: ANALYSIS OF A LOW-MELTING ALLOY. The analysis
of low-melting alloys such as Wood’s metal is greatly simplified by
complexometric titration, and tedious gravimetric separations are avoided. The
alloy is treated with concentrated nitric acid, evaporated to a small volume, and
after dilution the precipitated tin(I'V) oxide is filtered off; heavy metals adsorbed
by the precipitate are removed by washing with a known volume of standard
EDTA solution previously made slightly alkaline with aqueous ammonia. The
hydrated tin(I'V) oxide is ignited and weighed. The Bi, Pb, and Cd are determined
in the combined filtrate and washings from the tin separation ; these are diluted to
a known volume and aliquots used in the subsequent titrations. The Bi content is
determined by titration with standard EDTA at pH 1-2 using Xylenol Orange as
indicator; then, after adjustment of the pH to 5-6 with hexamine, the combined
Pb-hCd can be titrated with EDTA. 1,10-Phenanthroline may now be added to
mask the cadmium, and the liberated EDTA is titrated with standard lead nitrate
solution; this gives the cadmium content and thence the Pb content is obtained
by difference.

Procedure. Prepare a standard solution of lead nitrate {0.05M), a 0.05 per
cent aqueous solution of Xylenol Orange indicator, and a 1 ;10 -Phenanthroline
solution (O.OSAf) by dissolving 0.90 g of pure 1,10-phenanthroline in 1.5 cm^ of
concentrated nitric acid and 100 cm^ of water.

Weigh out accurately 2.0-2.5 g of Wood’s metal and dissolve it in hot
concentrated nitric acid (ca. 50 cm?). Evaporate the resulting solution to a small

X, 73 QUANTITATIVE INORGANIC ANALYSIS

volume, dilute to about 150cm^ with water and boil for 1-2 minutes. Filter offthc
precipitate of hvdrated tinflV) o.vidc through a quantitative filter paper
(Whatman No. 542) and keep the filtrate. Render a known volume (say 50.0 cm^)
of a 0.05Af-EDTA solution slightly basic with aqueous ammonia. Wash the
precipitate on the filter with this solution and then with 50 cm^ ofwatcr. The final
wash liquid should give no precipitate with 5 per cent sodium sulphide solution.
Iransfer the filtrate and washings (containing the other metals as w'cll as the
c.xccss of HDTA) quantitatively to a 500-cm’ graduated flask, and dilute to the
mark with dc-ionised water. Char the filter paper in the usual way and, after
ignition, weigh the tin(lV) oxide.

Into a conical flask, pipette a 50.0 or 100,0 cm-' aliquot of the solution and
adjust the pH to 1-2 with aqueous ammonia solution (use pH test paper). Add 5
drops of Xylcnol Orange indicator and titrate with additional 0,05A/-EDTA
until the colour changes .sharply from red to yellow. Tlris gives the fii content.
Record the total (combined) volume of EOT.A solution used. Now add small
amounts of hcxaniine (ca. 5 g) until an intense red-violet coloration persists, and
titrate with the standard EDT.A to a yellow ciui-point; the further consumption
of EDTA corresponds to the Pb-f C‘d content.

To determine ilic Cd content, add 20-25 cm-* of the l.IO-phcnanthrolinc
solution and titrate the liberated EDTA ss'ilh the 0.05Af-Iead nitrate solution
until the colour change from jellow' to red-violet occurs — a little practice is
required todisccrti the end-point precisely. Introduce funhet 2 -5-cm-' portionsof
the 1,10-phennnthrolincsolution and note whether the indicator colour ch.angcs:
if so. continue the titration with the lea.d nitrate solution, Tlic consumption of
lead nitrate solution corresponds to the Cd content.

Determination of anions

Anions do not complex directly with EDTA. but inclliods can be devised for
the determination ofappropriatc anions which involve either adding an exassof
a solution containing a cuion which reacts w itli the anion to be determined, and
then using EDTA to measure tiic excess of cation added; alternatively, the anion
is precipitated with a suitable cation, the precipitate is collected, dis'soivcd in
excess EDTA solution and then the excess EDTA is titrated with a standard
solution of an appropriate cation, Tltc procedure involved in the first method will
be self-evident but some ilctails arc given for determinations carried out by the
second method.

X,73. DETERMINATION OF HAlJDIuS (EXCI.UDING FLUORIDE)
AND TIIfOCVANAlES. Tlic procedure involved in the determination of
these anions is virtually that di.scusscd in Section X, 59 for the indirect
determination of silver. The anion to be determined is precipitateti as the silver
salt, the precipitate is collected and di.ssolvcd in a solution of potassium
tctracyanonickclatcfll) in the presence of an arnmonia/ammonium chloride
bulTer. Nickel ions arc liberated and titrated with standard EDTA solution using
murexide as indicator;

2Ag^ +[Ni(CN),)^- ^ Ni=* +2[Ag(CN);]-

Thc method may be illustrated by the determination of bromide; details for the
preparation of the potassium tctracyanonickclalc arc given in Section X, 59.

334

TITRIMETRIC ANALYSIS X, 74/75

Pipette 25.0 ,cm^ of the bromide ion_ solution (0.01-0.02M) into a 400-cm^
beaker, add excess, of dilute silver nitrate solution, filter off the precipitated silver
bromide on a sintered glass filtering crucible, and wash it with cold water.
Dissolve the precipitate in a warm solution prepared from 1 5 cm^ of concentrated
ammonia solution, 15 cm^ of IM-ammonium chloride, and 0.3 g of potassium
tetracyanonickelate. Dilute to 100-200-cm^, add 3 drops of murexide indicator,
and titrate with standard EDTA (.OIM) (slowly near, the end-point) until the
colour changes from yellow to violet. . , ,

1 mole EDTA = 2 moles Br~

X, 74. DETERMINATION OF PHOSPHATES, The . phosphate is
precipitated as Mg(NH 4 )P 04 , 6 H 20 , the precipitate is filtered off, washed,
dissolved in dilute hydrochloric acid, an excess of standard EDTA solution
added, the pH adjusted to 10, and the excess of EDTA titrated with standard
magnesium chloride or magnesium sulphate solution using Solochrome Black
(Eriochrome Black T) as indicator. The initial precipitation may be carried out in,
the presence of a variety of metals by first adding sufficient EDTA solution (IM)
to form complexes with all the polyvalent metal cations, then adding excess of
magnesium sulphate solution, followed by ammonia solution : alternatively, the
cations may be removed by passing the solution through a cation exchange resin
in the hydrogen form.

Procedure. Prepare a standard (0.05M) solution of magnesium sulphate or
chloride from pure magnesium (Section X, 62), an ammonia-ammonium chloride
bujfer solution (pH 10) (Section X, 55), and a standard (0.05M) solution of EDTA:

Pipette 25.0 cm^ of the phosphate solution (approx. 0.05M) into a 250 crn^
beaker and dilute to 50 cm^ with de-ionised water; add 1 cm^ of concentrated
hydrochloric acid and a few drops of methyl red indicator. Treat with an excess of
IM-magnesium sulphate solution (co. .2 cm^), heat the solution to boiling, and
add concentrated ammonia solution dropwise and with vigorous stirring until
the indicator turns yellow, followed by a further 2 cm^. Allow to stand for seyeral
hours or overnight. Filter the precipitate through a sintered-glass crucible
(porosity G4) and wash thoroughly with IM-ammonia solution (about 100 cm^).
Rinse the beaker (in which the precipitation was made) with 25 cm^ of hot IM-
hydrochloric acid and allow the liquid to percolate through the filter crucible,
thus dissolving the precipitate. Wash the beaker and crucible with a further 10
cm^ of IM-hydrochloric acid and then with about 75 cm^ of water. To the filtrate
and washings in the filter flask add 35.0 cm^ of 0.05M-EDTA, neutralise the
solution with IM-sodium hydroxide, add 4 cm^ of buffer solution and a few drops
of Solochrome Black (Eriochrome Black T) indicator. Back-titrate with standard
0.05M-magnesium chloride until the colour changes from blue to wine red. .

X, 75. DETERMINATION OF SULPHATES. The sulphate is precipitated
as barium sulphate from acid solution, the precipitate is filtered off and dissolved
in a measured excess of standard EDTA solution in the , presence of aqueous
ammonia. The excess of EDTA is then titrated with : standard magnesium
chloride solution using Solochrome Black (Eriochrome Black T) as.indicator.

Procedure. Prepare a standard magnesium chloride solution (0.05M) and a

uffer solution (pH 10); see previous Section. Standard EDTA (0.05M) will also be
required.

335

X. 76 QUANTITATIVB INORGANIC ANALYSIS

Pipette 25.0 cni^ of the sulplintc solution (0.02-0.03A/) into a 250-cm’ beaVer,
dilute to 50cm-\ and adjust thepH to I with 2A/-hydrochloric acid: heat nearly to
boiling. Add 1 5 cni^ of a nearly boiling barium chloride .solution (at. 0,05Af)fa'irIy
rapidly and with vigorous stirring: heat on a steam bath for 1 hour. Filter with
suction through a filter-paptfrdiscfWhatman filter paper No. 42) supported upon
a porcelain filter disc or a Gooch crucible, wash the precipitate thoroughly with
cold water, and drain. Tran.sfer the filter-paper disc and precipitate quantitatively
to the original beaker, add 35.0 cm-* standard 0.05A/-HDTA solution, 5 cin^
concentrated ammonia solution and boil gently for 1 5--20 minutes: add a further
2 cm' concentrated ammonia solution after 10-15 minutes to facilitate the
dis.solution of the precipitate. Cool the resulting clear solution, add 10 cm^ of the
buffer solution (pH = 10), a few drops of Solochromc Black (Eriochronic Black
T) indicator, and titrate the excess of FDTA witli the standard magnesium
chloride solution to a clear red colour.

Sulphate can also he determined by an c.xactly similar procedure by
precipitation as lead sulphate from a solution containing 50 per cent (by volume)
of propan-2-ol (to reduce the solubility of the lead sulphate), separation of the
precipitate, dissolution in excess ofstandard F DTA solution, and back-titration
of the excess FDTA with a standard 7inc solution using Solochrome Black
(Eriochrome Black T) as indicator.

B.3 PRECIPITATION TiTRATIONS

In the following Sections we are cona-rned with the use of standard solutions of
reagents such as silver nitrate, sodium chloride, potassium (or ammonium)
thiocyanate, and potassium cyanide. ,As already pointed out. some of the
determinations which will be considered strictly involve complex formation
rather than precipitation reactions, but it is convenient to group them here as
reactions involving the use of standard silver nitrate solutions. Before
commencing the experimental work, the theoretical Sections X, 29 and X, 30
should be studied.

X, 76. PREFARATfON OF O.lAf-SlLVF.R NITRATM Dhcusfion. Very
pure silver can be obtained commercially, and a standard solution can be
prepared by dissolving a known weight (say, i 0.787 g) in pure dilute nitric acid in
a conical flask having a funnel in the neck to prevent mechanical los.s, and making
uptoa known volume (say, I dm 'for a O.lAf solution). The presence of .acid must,
however, be avoided in determinations with potassium chromate as indicator or
in determinations employing adsorption indicators. It is therefore preferable to
employ a neutral solution prepared l>y dissolving silver nitrate (molecular weight,

169.87) in water.

A.R. Silver nitrate has a purity of at least 99.9 per t'cnt, so that a stand.ird
solution can be prepared by direct wcigliing. If, however, commercial re-
crystalliscd silver nitrate be employed, or if an addition.'tl check of the molarity of
the silver nitrate solution is required, standardisation may be cfl'cctcd with pure
sodium chloride. A.R. Sodium chloride has a purity of 9^9-100.0 per cent; the
substance is therefore an excellent primary standard. Sodium chloride is very
slightly hygroscopic, and for accurate work it is best to dry the finely powdered

336

TITRIMETRIC ANALYSIS X, 77

solid in an electric oven at 250-350 °C for 1-2 hours, and allow it to cool in a
desiccator.

• Procedure From A.R. silver nitrate. Dry some finely powdered A.R.
silver nitrate at 120 °C for 2 hours and allow it to cool in a covered vessel in a
desiccator. Weigh out accurately 8.494 g, dissolve it in water and make up to 500
cm^ in a graduated flask. This gives a O.IOOOM solution. Alternatively, about 8.5 g
of pure, dry silver nitrate may be weighed out accurately, dissolved in 500 cm^ of
water in a graduated flask, and the molar concentration calculated from the
weight of silver nitrate employed.

In many cases the A.R. material may be replaced by ‘pure recrystallised’ silver
nitrate, but in that case it is advisable to standardise the solution against sodium
chloride. Solutions of silver nitrate should be protected from light and are best
stored in amber-coloured glass bottles.

X, 77. STANDARDISATION OF THE SILVER NITRATE SOLUTION.
Sodium chloride has a molecular weight of 58.44. A O.IOOOM solution is prepared
by weighing out 2.922 g of the pure dry A.R. salt (see Section X, 76) and dissolving
it in 500 cm^ of water in a graduated flask. Alternatively about 2.9 g of the pure
salt is accurately weighed out, dissolved in 500 cm^ of water in a graduated flask,
and the molar concentration calculated from the weight of sodium chloride
employed.

A. With potassium chromate as indicator. The Mohr titration. The
reader is referred to Section X, 30 for the detailed theory of the titration. Prepare
the indicator solution by dissolving 5 g A.R. potassium chromate in 100 cm^ of
water. The final volume of the solution in the titration is 50-100 cm^, and 1 cm^ of
the indicator solution is used, so that the indicator concentration in the actual
titration is 0.005-0.0025M.

Alternatively, and preferably, dissolve 4.2 g A.R. potassium chromate and 0.7 g
A.R. potassium dichromate in 100 cm^ of water; use 1 cm^ of indicator solution
for each 50 cm^ of the final volume of the test solution.

Pipette 25 cm^ of the standard O.lM-sodium chloride into a 250-cm^ conical
flask resting upon a white tile (1), and add 1 cm^ of the indicator solution
(preferably with a 1-cm^ pipette). Add the silver nitrate solution slowly from a
burette, swirling the liquid constantly, until the red colour formed by the addition
of each drop begins to disappear more slowly: this is an indication that most of
the chloride has been precipitated. Continue the addition dropwise until a faint
but distinct change in colour occurs. This faint reddish-brown colour should
persist after brisk shaking. If the end-point is overstepped (production of a deep
reddish-brown colour), add more of the chloride solution and titrate again.
Determine the indicator blank correction by adding 1 cm^ of the indicator to a
volume of water equal to the final volume in the titration (2), and then O.OIM-
silver nitrate solution until the colour of the blank matches that of the solution
titrated. The indicator blank correction, which should not amount to more than
0.03-0.10 cm^ of silver nitrate, is deducted from the volume of silver nitrate used
in the titration. Repeat the titration with two further 25-cm^ portions of the
sodium chloride solution. The various titrations should agree within 0.1 cm^.

Notes. 1. The end-point is very readily detected in a large porcelain basin.
1 he solution is stirred with a short glass stirring rod.

2. A better blank is obtained by adding about 0.5 g of A.R. calcium carbonate

etore determining the correction. This gives an inert white precipitate similar to

337

X, 78 QUANTITATIVE INORGANIC ANALYSIS

tliat obtained in the titration of cltlorides and materially assists in matching the
colour tints.

B. \\'illi an adsorption indicator. Discussion. Tiic detailed theory of the
process is given in Section X. 30. Both lltiorcsccin and dichlorofluorescein are
suitable for the titration of cltlorides. In both cases the end-point is reached when
the white precipitate in the preentsh-yeilow solution suddenly assumes a
pronounced reddish tint. The change is reversible upon the addition of chloride.
With fluorescein the solution nnist be neutral or only faintly acidic with acetic
acid; acid solutions should l>c treated wiili a slight exce.ss of sodium acetate. Tltc
chloride solution should be diluted to about 0.01 -0.05 A/, for if it Is more
concentrated the precipitate coagulnle.s too soon and interferes, Fluorescein
cannot be used in solutions more dilute than 0.005 .\f. Willi more dilute .solutions
resort must be made to dichlorofluorescein. which possesses other advantages
over fluorescein. Dichlorofluorescein gives good results in very dilute solutions
(c.g.. for drinking water) and applicable in the presence of acetic acid and in
weakly acid solutions. For this reason thcchloridc.s of copper, nickel, manganese,
zinc, aluminium, and magnesium, which cannot be titrated according to the
mctliod of Mohr, can be determined by a direct titration when
dichlorofluorescein is used as indicator.

For the reverse titration (chloride into siK er nit rate), lartrazinc (4 drops of a 0.2
per cent .solution per HK) cm ') is a good indicator. At the end-point, the almost
colourless liquid assumes a blue colour.

Tlie indicator solutions arc prepared as follows;

I'liiorescciiu Dissohe 0.2 g fluorescein in llKl cm 'of 70 per cent ethanol, or
dissolve 0.2 g, sodium fliioresceinate in 100 env' of water.

Oichlorofluorcsceia DissolvcO.l gdichlorofluore.sceinin lOOcm^ of60-70
per cent ethanol, or dis'^olvc O.l g sodium dich!orofluorc,sa:inate in 100 enr' of
water.

Procedure. Pipette 25 cm' of the standard 0.1 A/-sodium chloride into a
250-cm^ conical flask. Add U) drops of either fluorescein or dicfilorofluore.sa'in
indicator, and titrate with the silver nitrate solution in a diffuse light, while
rotating the flask constantly. .As the end point is approached, the silver chloride
coagulates appreciably, and the Uval development of a pink colour upon the
addition of a drop of the silver nitrate solution becomes more and more
pronounced. Continue the addition of the silver nitrate .solution until the
precipitate suddenly a.ssumc.s a pronounced pink or red colour. Repeat the
titration with two other 25-cm' portions of the chloride solution. Individual
titrations should agree within 0.1 cm\

Calculate the molar concentration of the silver nitrate solution.

X, 78. DETER.MFNATION OF CIIEORI Dm Cither the Mohr titration or
the adsorption-indicator method may be used for the determination of chlorides
in neutral solution by titration with standard 0. 1 A/-siIvcr nitrate. If the solution is
acid, neutralisation may be effected with cliloridc-frec calcium carbonate, sodium
(eiraboraie, or sodium hydrogen carbonate; the A.R. substances are suitable.
Mineral acid may also be removed by neutralisinc most of the acid with ammonia
solution ami then adding an c.xccss of A.R, ammonium acetate. Titration of the
neutral solution, prepared with calcium carbonate, by the adsorption indicator
method is rendered easier by tlic addition of 5 enr' of 2 percent dc.xtrin solution;
this offsets the coagulating cflcct of the calcium ion. If the solution is ba.sic, it may

338

TITRIMETRIC ANALYSIS X, 79/80

be neutralised with chloride-free nitric acid, using phenolphthalein as indicator.

Similar remarks apply to the determination of bromides ; the Mohr titration can
be used, and the most suitable adsorption indicator is eosin which can be used in
dilute solutions and even in the presence of O.IM nitric acid, but in general, acetic
acid solutions are preferred. Fluorescein may be used but is subject to the same
limitatioiis as experienced with chlorides (Section X, 77, B). With eosin indicator,
the silver bromide flocculates approximately 1 per cent before the equivalence
point and the local development of a red colour becomes more and more
pronounced with the addition of silver nitrate solution; at the end-point the
precipitate assumes a magenta colour.

The indicator is prepared by dissolving 0.1 g eosin in 100 cm^ of 70 per cent
ethanol, or by dissolving 0.1 g of the sodium salt in 100 cm^ of water.

For the reverse titration (bromide into silver nitrate), rhodamine 6G (10 drops
of a 0.05 per cent aqueous solution) is an excellent indicator. The solution is best
adjusted to 0.05M with respect to silver ion. The precipitate acquires a violet
colour at the end-point.

Thiocyanates may also be determined using adsorption indicators in exactly
similar manner to chlorides and bromides, but an iron(III) salt indicator is
usually preferred (Section X, 82).

X,79. DETERMINATION OF IODIDES. Discussion. The Mohr method
cannot be applied to the titration of iodides (or of thiocyanates), because of
adsorption phenomena and the difficulty of distinguishing the colour change of
the potassium chromate. Eosin is a suitable adsorption indicator, but di-
iododimethylfluorescein is better. Eosin is employed as described under
bromides (Section X, 78).

The di-iododimethylfluorescein indicator is prepared by dissolving 1.0 g in 100
cm^ of 70 per cent ethanol. The colour change is from an orange-red to a blue-red
on the precipitate.

X,80. DETERMINATION OF MIXTURES OF HALIDES WITH
ADSORPTION INDICATORS. A. Chloride and iodide in admixture. These
two ions differ considerably in the ease with which they are adsorbed on the
corresponding silver halide. This makes it possible to select adsorption indicators
which will permit the determination of chloride and iodide in the presence of
one another. Thus the iodide may be determined by titration with standard O.IM-
silver nitrate using di-iododimethylfluorescein and the iodide + chloride by a
similar titration using fluorescein. Chloride is obtained by difference. If a large
excess of chloride is present the result for iodide may be as much as 1 per cent
high. If, however, Rose Bengal (dichlorotetraiodofluorescein) is used as indicator
(colour change, carmine red to blue-red) in the presence of ammonium carbonate,
the iodide titration is exact.

B. Bromide and iodide in admixture. The total halide (bromide -t- iodide) is
determined by titration with standard O.lM-silver nitrate - using eosin or
fluorescein as indicator. The iodide is determined by titration with 0.01-0.2M-
silver nitrate, using di-iododimethylfluorescein as indicator. Bromide is obtained
by difference.

Numerous adsorption indicators have been suggested for various purposes,

ut a full treatment is outside the scope of this work.

339

X, 81/82 QUANTITATIVI- INORGANIC ANALYSIS

X, 81. DETERMINATION OF MIXTURI^S OF HALIDES BY AN
INDIRECI' METHOD. Disiitssion. The inctliod is nppliaiblc lo the
determination of a mixture of two salts having llic same anion (c.g., sodium and
potassium chlorides) or the same cation (e.g,. potassium chloride and potassium
bromide). Let us first suppose that it is desired to determine the amount of sodium
and potassium chlorides in a mixture of the two salts, A known weight g)of
the solid mixture is taken, and the total chloride is determined with .standard
0.1 A/-silvcr nitrate, using Mohr’s method or an adsorption indicator. Let Wj gof
silver nitrate be required for tlic complete precipitation of g of the mixture,
which contains .v g of NaC! and y g of KCI. Then;

I69.87.V ]69.87p _

T8l4'"^"''74T5 "*

Upon solving these two sintultancous equations, the values for x and y arc
deduced.

Let us now suppose that the determination of potassium chloride and
potassium bromide in admixture is desired. Tiic total halide is determined by
Mohr's method or with an adsorption indicator. Let tlie weight of the mixture be
"3 g- "'j c the weight of silver nitrate required for complete precipitation, pg be
the weight of the pot.assium chloride, and q g be the weight of the potassium
bromide. Then:

p-yq « Wj

m^p^l69.87</ _

74.55"'^ "il9T)0' ’'■*

The values of p and q can be obtained by solving the simultaneous equations.

It can be shown that the method depends upon the difference lietwecn the
molecular weights of the two components of the mixture and that inicr alia it is
most satisfactory when the two constituents are pre.senl in approximately equal
proportions.

X,81 PREPARATION AND USE OF O.IM-AMMONIUM OR
POTASSIUM THIOCVAN.ATE. Titrations according to Volliard's method.
Discussinn. Volhard's original method for the determination of silver in ilihdc
nitric aciil solution by titration with si.intlard tliiocyanatc solution in the
presence of an iron(lll) sail as indicator has proved of great value not only for
silver determinations, but also in numerous indirect analyses. The theory of the
Voliiard process has been given in Section 30. In this'conncclion it must be
pointed out that the concentration of the nitric acid should be from 0.5-LSAf
(strong nitric acid retards the formation of the tliiocyanatoironflll) complex
[FeSCN]' ^ ) and at a temperature not c.xcceding 25 ‘C (higher tcmpcraturc.s tend
to bleach the colour of the indictitor). The solutions must be free from nitroms
acid, which gives a red colour with tliiocyanic acid, and may be mistaken for
‘irontlll) thiocyanate’. Pure nitric acid is prepared by diluting the usual pure (c.g.,
A.R.jacid with about one-fourth of its volume of water and boiling until perfectly
colourless; this eliminates any lower oxides of nitrogen which may be present.

The method may be applied to those anions (e.g., chloride, bromide, and

340

TITRIMETRIC ANALYSIS X, 82

iodide) which are completely precipitated by silver and are sparingly soluble in
dilute nitric, acid. Excess of standard silver nitrate solution is added to the
solution containing free nitric acid, and the residual silver nitrate solution is
titrated .with standard thiocyanate solution. This is sometimes termed the
residual process. Anions whose silver salts are slightly soluble in water, but which
are soluble in nitric acid, such as phosphate, arsenate, chromate, sulphide, and
oxalate, may be precipitated in neutral solution with an excess of standard silver
nitrate solution. The precipitate is filtered off, thoroughly washed, dissolved in
dilute nitric acid, and the silver titrated with thiocyanate solution. Alternatively,
the residual silver nitrate in the filtrate from the precipitation may be determined
with thiocyanate solution after acidification with dilute nitric acid.

Both ammonium and potassium thiocyanates are usually available as
deliquescent solids; the A.R. products are, however, free from chlorides and other
interfering substances. An approximately O.IM solution is therefore first
prepared, and this is standardised by titration against standard O.lM-silver
nitrate.

Procedure. Weigh out about 8.5 g A.R. ammonium thiocyanate, or 10.5 g
A.R. potassium thiocyanate, and dissolve it in 1 litre of water in a graduated flask.
Shake well.

Standardisation. Use O.lM-silver nitrate, which has been prepared and
standardised as described in Section X, 77.

The Iron(in) indicator solution consists of a cold, saturated solution of A.R.
ammonium iron(III) sulphate in water (about 40 per cent) to which a few drops of
6M-nitric acid has been added. One cm^ of this solution is employed for each
titration.

Pipette 25 cm^ of the standard O.lM-silver nitrate into a 250 cm^ conical flask,
add 5 cm^ of 6M-nitric acid and 1 cm^ of the iron(III) indicator solution. Run in
the potassium or ammonium thiocyanate solution from a burette. At first a white
precipitate is produced, rendering the liquid of a milky appearance, and as each
drop of thiocyanate falls in, it produces a reddish-brown cloud, which quickly
disappears on shaking. As the end-point approaches, the precipitate becomes
flocculent and settles easily; finally one drop of the thiocyanate solution produces
a faint brown colour, which no longer disappears upon shaking. This is the end-
point. The indicator blank amounts to 0.01 cm^ of O.lM-silver nitrate. It is
essential to shake vigorously during the titration in order to obtain correct
results.*

The standard solution thus prepared is stable for a very long period , if
evaporation is prevented.

Use of tartrazine as indicator. Satisfactory results may be obtained by the
use of tartrazine as indicator. Proceed as above, but add 4 drops of tartrazine (0.5
per cent aqueous solution) in lieu of the iron(III) indicator. The precipitate will
appear pale yellow during the titration, but the supernatant liquid (best viewed
by placing the eye at the level of the liquid and looking through it) is colourless.
At the end-point, the supernatant liquid assumes a bright lemon-yellow colour,
fhe titration is sharp to one drop of 0. IM-thiocyanate solution. ■

freshly precipitated silver thiocyanate adsorbs silver ions, thereby causing a false end-point
icti, however, disappears with vigorous shaking.

341

X, 83/84 QUANTITATIVE INORGANIC ANALYSIS

X,83. DETERMINATION OF SILVER IN A SILVER ALLOY. A
commercial silver alloy in the form of wire or foil is suimblc for this
determination. Clean the alloy witli emery cloth and weigh it accurately. Place it
in a 250-cm^ conical flask, add 5 cm’ water and 10 cm’ concentrated nitric acid;
place a funnel in the mouth of the flask to avoid mechanical loss. Warm the flask
gently until the alloy has dissolved. Add a little water and boil for 5 minutes in
order to c.xpcl o.sidcs of nitrogen. Transfer the cold solution quantitatively to a
100-cm’ standard flask and make up to the mark with distilled water. Titrate 25-
cm’ portions of the solution with standard 0. 1 ,\f-thiocyanatc.

1 mole KSCN h 1 mole .Ag *

Note. The presence of metals who.se salts arc colourless does not influence
the accuracy of the delerminaiion, except that mercury and palladium must
absent .since their thiocyannics arc insoluble. .Salts of metaks (c.g^ nickel and
cobalt) which arc coloured nuisl not be present to any considerable c.xlcnL
Copper docs not interfere, provided it does not form more than about 40 percent
of the alloy.

X. 84. DETERMINATION OF CHLORIDES (VOLHARD’S
METHOD). Disi'iisiion. The chloride solution is treated with c.tccss of
standard silver nitrate solution, and the residual .silver nitrate determined by
titration with standard thiocyanate solution. Now silver chloride is more .soluble
than silver thiocviinatc. and would react with the thiocyanate thiLs:

AgCl (solid) -t-SCN' :e=AcSCN Isolid) -f-C) '

It is therefore necessary to remove the silver chloride by filtration. The filtration
may be avoided by the addition of a little nil robenrenc (about 1 cm’ for each 0.05
g of chloride); the silver chloride particles are prob.ably surrounded by a film of
nitrobenzene. Another method, applicable to chlorides, in which filtration of the
silver chloride is unnecessary, is to employ tartraz.ine as indicator (Section X. 82).

Procedure A (HCl content of concentrated hydrochloric acid). Ordinary
concentrated hydrochloric acid is usually K)~1IaV. and must be diluted first.
Measure out accurately 10 cm’ of the* concentrated acid from a burclte into a I-
dm’ graduated flask and make up to the mark with distilled water. Shake well.
Pipette 25 cm’ into a 250.cnr' conical flask, add 5 cm’ C>,M-nitric acid and then
add 30 cm’ standard 0. !A/-silvcr nitrate for .sulTicient to give 2''5 cm’ cxcc.ss).
Sliake to coagulate the precipitate.’ filter through a quantitative filter paper (or
through a porous porcelain or sintcrcd-gla.ss crucihic), ami wash thoroughly with
very dilute nitric acid (1 ;1(>0). Add 1 cm’ of iheiron(in) indicator solution to the
combined filtrate and washings, and titrate the residual silver nitrate with
standard O.lAf-iliiocyanatc.

Calculate the volume of standard 0. 1 Af-silver nitrate that lias reacted with the
hydrocliloric acid, and therefrom the percentage of I !Ci in the sample employed.

Procedure B. Pipette 25 cm’ of the diluted solution into a 250-cm’ conical
fla.sk containing 5 cm’ 6A/-nitricacid. Add a slight cxcc.ss of standard 0. 1 A/-silvcr

‘Ills heller 10 boil ihc suspension for n feiv minutes to co.ijiuKite the sitter chloride .nnd thus remote
most of the adsorbed silver ions from il.s surface before filtration.

.342

TITRIMETRIC ANALYSIS X, 85

nitrate (about 30 cm^ in all) from a burette. Then add 2-3 cm^ pure (e.g., A.R.)
nitrobenzene and 1 cm^ of the iron(III) indicator, and shake vigorously to
coagulate the precipitate. Titrate the residual silver nitrate with standard O.IM-
thiocyanate until a permanent faint reddish-brown coloration appears.

From the volume of silver nitrate solution added, subtract the volume of silver
nitrate solution that is equivalent to the volume of standard thiocyanate required.
Then calculate the percentage of HCl in the sample.

Procedure C. Pipette 25 cm^ of the diluted solution into a 250-cm^ conical
flask containing 5 cm^ of 6M-nitric acid, add a slight excess of O.lM-silver nitrate
(30-35 cm^) from' a burette, and 4 drops of tartrazine indicator (0.5 per cent
aqueous solution). Shake the suspension for about a minute in order to ensure
that the indicator is adsorbed on the precipitate as far as possible. Titrate the
residual silver nitrate with standard O.lM-ammonium or potassium thiocyanate
with swirling of the suspension until the very pale yellow supernatant liquid
(viewed with the eye at the level of the liquid) assumes a rich lemon-yellow colour.

Bromides can likewise be determined by the Volhard method, but as silver
bromide is less soluble than silver thiocyanate it is not necessary to filter off the
silver bromide (compare chloride). The bromide solution is acidified with dilute
nitric acid, an excess of standard O.lM-silver nitrate added, the mixture
thoroughly shaken, and the residual silver nitrate determined with standard
O.lM-ammonium or potassium thiocyanate, using iron(III) alum as indicator. •

Iodides can also be determined by this method, and in this case too there is no
need to filter off the silver halide, since silver iodide is very much less soluble than
silver thiocyanate. In this determination the iodide solution must be very dilute in
order to reduce adsorption effects. The dilute iodide solution (ca. 300 cm^),
acidified with dilute nitric acid, is treated very slowly and with vigorous stirring
or shaking with standard O.lM-silver nitrate until the yellow precipitate
coagulates and the supernatant liquid appears colourless. Silver nitrate is then
present in excess. One cm^ of iron(III) alum solution is added, and the residual
silver nitrate is titrated with standard O.lM-ammonium or potassium
thiocyanate.

X, 85. DETERMINATION OF FLUORIDE; PRECIPITATION AS LEAD
CHLOROFLUORIDE COUPLED WITH VOLHARD TITRATION. Dis-
cussion. This method is based upon the precipitation, of lead chlorofluoride, in
which the chlorine is determined by Volhard’s method, and from this result the
fluorine content can be calculated. The advantages of the method are: the
precipitate is granular, settles readily, and is easily filtered; the factor for
conversion to fluorine is low; the procedure is carried out at pH 3.6-5.6, so that
substances which might be co-precipitated, such as phosphates, sulphates,
chromates, and carbonates, do not interfere. Aluminium must be entirely absent,
since even very small quantities cause low results; a similar effect is produced by
boron (>0.05 g), ammonium (> 0.5 g), and sodium or potassium (> 10 g) in the
presence of about 0.1 g of fluoride. Iron must be removed, but zinc is without
effect. Silica does not vitiate the method, but causes difficulties in filtration. .

Procedure. Pipette 25.0 cm^ of the solution containing between 0.01-0.1 g
uoride into a 400-cm^ beaker, add 2 drops of bromo-phenol blue indicator, 3
cm of 10 per cent sodium chloride, and dilute the mixture to 250 cm^. Add dilute
nitnc acid until the colour just changes to yellow, and then add dilute sodium
ydroxide solution until the colour just changes to blue. Treat with 1 cm^ of

343

X, 86 QUANTITATIVE INORGANIC ANALYSIS

conccnirated hydrocliloric :\cid, then with 5,0 g of A.R. lead nitrate, and lieat on
tlie .steam bath. Stir gently until the lead nitrate has dissolved, and then
immediately add 5.0 g crystallised sodium acetate ;ind stir vigorously. Digest on
the steam batli for 30 minutes, with occasional stirring, and allow to stand
overnight.

Meanwhile, a wasliing solution of lead chloronuoridc is prepared as follows.
Add a solution of 10 g of lead nitrate in 300 cm' of water to 100 of a solution

containing 1.0 g of sodium nuorideand 2cm^ of concentrated hydrochloric acid,
mix it thoroughly, and allow the precipitate to settle. Decant the supernatant
liquid, wash the precipitate by decantation with 5 portions of water, each ofabout
200 cnv’. Finally add ! Hire of water to the precipitate, shake the mixture at
intervals during an hour, allow the precipitate to settle, and filter the liquid.
Further quantities of wash liquid may be prepared as needed by treating the
precipitate with fre.sli portions of water. The .solubility of lead chlorotluoridc in
water is 0.325 g dm ’ at 25 C.

.Separate the original precipitate by decantation through a Whatman No. 542
or No. 42 paper. Transfer the precipitate to the filter, wash once with cold water,
four or five times with the saturated solution of lead chloroffuoridc.and finally
once more with cold water. Transfer the precipitate and paper to the beaker in
which precipitation was made, stir the paper to n pulp in KX) cm^ of 5 per cent
nitric acid, and heat on the steam b.ath until the precipitate has dissolved (5
mimitcsi. Add a slight excess of standard 0. 1 .\/-.silvcr nitrate, digest on the steam
bath for a further 30 minutes, and allow to cool to room temperature while
protected from the light. Filter the precipitate ofsilvcr chloride through a sintered
glass crucihic, wash with a lit tie cold water, and titrate the residual silver nitrate in
the filtrate and washings with standard 0,1 ,\f-thioc\anate. Subtract the amount
ofsilvcr found in the filtrate from that originally added. The difference represents
the amount of silver that was required to combine with the chlorine in the lead
chlorofiiioride precipitate.

I Mole AgNOj r: 1 Mole 1- '

X, 86. DETFRMINATION OF AR.SF.,\ATMS. Phciismn. Arsenale.s in
.solution arc precipitated as silver ar.scnaie. Ag vAsO^, by the addition of neutral
silver nitrate solution; the solution must be neutral, or if slightly acid, an excess of
sodium acetate must lie present to reduce the acidity; if .strongly acid, most of the
acid should be neutralised by aqueous sodium hydroxide. Tlic silver arsenate is
dissolved in dilute nitric acid, and the .sih'er titrated with standard thiocyanate
solution. The silver arsenate has nearly six limes the weight of the arsenic, hence
quite small amounts of arsenic may Ik’ determined by this procedure.

Arsenites may al.so be determined by this procedure but must first be oxidised
by treatment with nitric acid. Small amounts of antimony and tin do not interfere,
but chromates, phosphates, molybdates, tungstates, and vanadates, which
precipitate as the silver salts, should be absent. An excessix'c amount of
ammonium salts has a solvent action on the silver arsenate.

Procedure. Place 25 cm"' of the ar.scnate solution in a 250-cin’' beaker, add
an equal x'olurnc of distilled water and a few drops ofphcnolphthalein solution.
Add sullicicnl sodium hydroxide solution to give an alkaline reaction, and then
discharge the red colour from the solution by just acidifying xvith acetic acid. Add
a slight c.xcess of silver nitrate solution with vicorous stirring, and allow the

344

TITRIMETRIC ANALYSIS X,87

precipitate to settle in the dark. Pour off the supernatant liquid through a sintered
glass crucible, wash the precipitate by decantation with cold distilled water,
transfer the precipitate to the crucible, and wash it free from silver nitrate
solution. Wash out the receiver thoroughly. Dissolve the silver arsenate in dilute
nitric acid (ca. M) (which leaves any silver chloride undissolved}, wash with very
dilute nitric acid, and make up the filtrate and washings to 250 cm^ in a graduated
flask. Titrate a convenient aliquot portion with standard ammonium (or
potassium) thiocyanate solution in the presence of iron(III) alum as indicator.

3 moles KSCN = 1 mole As 04 ^“

X, 87. DETERMINATION OF CYANIDES. Discussion. The theory of
the titration of cyanides with silver nitrate solution has been given in Section X,
22. All silver salts except the sulphide are readily soluble in excess of a solution of
an alkali cyanide, hence chloride, bromide, and iodide do not interfere. The only
difficulty in obtaining a sharp end-point lies in the fact that silver cyanide is often
precipitated in a curdy form which does not readily re-dissolve, and, moreover,
the end-point is not easy to detect with accuracy.

There are two methods for overcoming these disadvantages. In the first the
precipitation of silver cyanoargentate at the end-point can be avoided by the
addition of ammonia solution, in which it is readily soluble and if a little
potassium iodide solution is added before the titration is commenced, sparingly
soluble silver iodide, which is insoluble in ammonia solution, will be precipitated
at the end point. The precipitation is best seen by viewing against a black
background.

In the second method diphenylcarbazide is employed as an adsorption
indicator. The end-point is marked by the pink colour becoming pale violet
(amost colourless) on the colloidal precipitate in dilute solution {ca. O.OIM) before
the opalescence is visible. In O.IM solutions, the colour change is observed on the
precipitated particles of silver cyanoargentate.

Procedure. NOTE. Potassium cyanide and all other cyanides are deadly
poisons, and extreme care must be taken in their use. Details for the disposal of
cyanides and other dangerous and toxic chemicals may be found in Refs 22 and 23.

For practice in the method, the cyanide content of potassium cyanide
(laboratory reagent grade) may be determined.

Method A. Weigh out accurately about 3.5 g potassium cyanide from a
glass-stoppered weighing bottle, dissolve it in water and make up to 250 cm^ in a
graduated flask. Shake well. Transfer 25.0 cm^ of this solution by means of a
burette and NOT a pipette to a 250-cm^ conical flask, add 75 cm^ water, 5-6 cm^
oM-ammonia solution, and 2 cm^ 10 per cent potassium iodide solution. Place
AH ^ paper, and titrate with standard O.lM-silver nitrate.

Add the silver nitrate solution dropwise as soon as the yellow colour of silver
iodide shows any signs of persisting. When 1 drop produces a permanent
turbidity, the end-point has been reached.

Method B. Prepare the solution and transfer 25 cm^ of it to a 250-cm^
conical flask as detailed under Method A. Add 2 to 3 drops of diphenylcarbazide
indicator and titrate with standard O.lM-silver nitrate solution until a permanent
violet colour is just produced.

345

X, 88 QUANTITATIVE INORGANIC ANALYSIS

The diphenylcarbazicle indicator is prepared by dissolving O.I g of the solid in
100 enr’ of ethanol.

1 molcAgNOj " 2 moles CN’.

X,88. DKIERMINATION OF CHLORIDES BY TITRATION WITH
.MERCURY(II) NITRATE SOLUTION. Dbaisshm. When chloride ions
arc titrated with mcrcurjHD ions (a solution of niercuryfl!) nitrate acidified
with nitric acid), the reaction
Hg'M-2CreiHpClj

is essentially stoichiometric. The reaction is not strictly a precipitation reaction,
hut it is convenient to include this alternative method of determining chlorides
with the argcntimetric method of Section X, 84. The end-point may be detected
with diphcnylcarbazone. which forms a blue-violet complex with mcrcurytll)
ions. Alternatively, a mixture of diphcnylcarbazone and bromophenol blue may
be used; the bromophenol bhieehanges from bluelalkaline) to yellow (acid) at ra.
pH 3,6. which is the acidity sometimes recommended for the titration in dilute
solutions. At the equivalence point the yellow colour of the solution bccomc-s
blue-violet owing to the reaction of the cxcc.ss of mercury(II) ions with
diphcnylcarbazone. Satisfactory results can. however, be obtained in the pH
range 3-8.

The mercurimetric method may be applied to the titration of chlorides in very
dilute solutions -down to 0- l(K) p.p.m. range. Bromides, thiocyanates, and
cyanides may be delennincd by similar methods, but there is no particular
advantage over the usual silver procedures. .Sodium, potassium, calcium,
magnesium, aluminium, mangane.se, zinc, fluoride, .sulphate, nitrate, and .acetate
individually in concentrations at least cqu.al to that of the chloride do not
interfere: chromate and iron(ni)ions read with diphenykarbazoncand must be
removed if present.

The main advantage of the mercurimetric method of determining chloride, i.c.,
its applicability to very dilute solutions of chloride, is only realised to the
maximum by working in an 80 per cent cthanolic medium: in purely aqueous
solution the end-point is not very sharp. -As the reaction is not strictly
stoichiometric, it is necessary to standardise the mcrcurydl) nitrate reagent
against sodium (or potassium) chloride.

Procedure. Prepare a mcrvur\ill] nitruic sn/iifiou (0.02»V) bv dissolving 3.4 g
rccry.slalliscd mercury(ll) nitrate IlgtNOUi.lLO, in 800 cm^ distilled water
containing 20 cm-* 2M nitric acid. Dilute to 1 litre in a graduated (lask and then
,standardi.sc with A.R, sodium chloride as described below. Prepare the indicator
by dissolving 0.1 g diphcnylcarbazone in 100 cm' of ethanol, and the mixed
indicator by dissolving 0,5 g diphenylcarbazonc and 0.5 g bromophenol blue in
100 cm^ of 95 percent ethanol.

To .'itandnrdise the nwrcury^II) nitrate solution, weich accurately about 1.5 g
A.R. sodium chloride and dissolve in I dm^ distilled water in a gr.aduatcd flask.
To 25.0 cm ' of this solution add I cm-’ of the diphcnylcarbazone indicator, and
titrate with the 0.02iV mcrcury(ll) nitrate solution until the first permanent blue-
purple coloration appears. Repeat the titration with a fresh portion of the
solution but using the mixed indicator, and decide which gives the most easily
recognisable change at the end-point. At this concentration of chloride the end-
point .should be quite sharp in purely aqueous solution, but at lower

346

TITRIMETRIC ANALYSIS X, 89

concentrations about 80 cm^ of 95 per cent ethanol should be added before
starting the titration.

The determination of the chloride content of a given solution will be apparent
from the above standardisation details.

X, 89. DETERMINATION OF POTASSIUM. Discussion. Potassium
may be precipitated with excess of sodium tetraphenylborate solution as
potassium tetraphenylborate. The excess of reagent is determined by titration
with mercury(II) nitrate solution. The indicator consists of a mixture of iron(III)
nitrate and dilute sodium thiocyanate solution. The end-point is revealed by the
decolorisation of the iron(III)-thiocyanate complex due to the formation of the

colourless mercury(II) thiocyanate. The reaction between mercury(II) nitrate and
sodium tetraphenylborate under the experimental conditions used is not quite
stoichiometric, hence it is necessary to determine a factor (cm^ of Hg(N 03)2
solution equivalent to 1 cm^ of NaB(C 6 H 5)4 solution). Halides must be alDsent.

Procedure. Prepare the sodium tetraphenylborate solution by dissolving 6.0 g
of the solid in about 200 cm^ of distilled water in a glass-stoppered bottle. Add
about 1 g of moist aluminium hydroxide gel, and shake well at five minute
intervals for about twenty minutes. Filter through a Whatman No. 40 filter paper,
pouring the first runnings back through the filter if necessary, to ensure a clear
filtrate. Add 15 cm^ of O.lM-sodium hydroxide to the solution to give a pH of
about 9, then make up to one litre and store the solution in a polythene bottle.
Prepare the mercury{II) nitrate solution as in the last Section, but using a weight of
approximately 10 g in a litre of solution; this solution may be standardised by
titrating with a standard thiocyanate solution using iron(III) alum as indicator.
Prepare the indicator solutions for the main titration by dissolving separately 5 g
hydrated iron(III) nitrate in 100 cm^ of distilled water and filtering, and 0.08 g
sodium thiocyanate in 100 cm^ distilled water.

Standardisation. Pipette 10.0 cm^ of the sodium tetraphenylborate solution
into a 250 cm^ beaker and add 90 cm^ water, 2.5 cm^ O.lM-nitric acid, 1.0 cm^
iron(III) nitrate solution, and 1.0 cm^ sodium thiocyanate solution. Without
delay stir the solution mechanically, then slowly add from a burette 10 drops of
niercury(II) nitrate solution. Continue the titration by adding the mercury(II)
nitrate solution at a rate of 1-2 drops per second until the colour of the indicator
IS temporarily discharged. Continue the titration more slowly, but maintain the
rapid rate of stirring. The end-point is arbitrarily defined as the point when the
indicator colour is discharged and fails to reappear for 1 minute. Perform at least
three titrations, and calculate the mean volume of mercury(II) nitrate solution
equivalent to 10.0 cm^ of the sodium tetraphenylborate solution.

Pipette 25.0 cm^ of the potassium ion solution (about 10 mg K"^) into a 50-cm^
graduated flask, add 0.5 cm^ M-nitric acid and mix. Introduce 20.0 cm^ of the
sodium tetraphenylborate solution, dilute to the mark, mix, then pour the
mixture into a 150-cm^ flask provided with a ground stopper. Shake the
stoppered flask for 5 minutes on a mechanical shaker to coagulate the precipitate,
then filter most of the solution through a dry Whatman No. 40 filter paper into a
ry beaker. Transfer 25.0 cm^ of the filtrate into a 250-cm^ conical flask and add
5 cm of water, 1.0 cm^ of iron(III) nitrate solution, and 1.0 cm^ of sodium

locyanate solution. Titrate with the mercury(II) nitrate solution as described
above.

347

X, 90 QUANTITATIVE INORGANIC ANALYSIS

B. 4 OXIDATION - REDUCTION TITRATIONS

In the following Sections we arc concerned with the titration of reducing agents
with oxidising agents such ns potassium permanganate, potassium dichromatc,
ccrium(IV) sulphate, iodine, potassium iodutc and potassium bromatc, and with
the titration of oxidising agents by reducing reagents such as arscnicfJU) oxide
and sodium thiosulphate.

The relevant theoretical Sections (X. 31-33) should be studied, and it should
also be noted that in many cases, before titration v\‘itii an oxidising reagent is
carried out, it is necessary to ensure that the substance to be titrated is in a
suitable lower oxidation state, i.c., it may be necessary to reduce the test solution
before titration can be carried out. ,A selection of methods for carrying out such
reductions is given at the end of this Chapter (Sections' X, 142-6).

Oxidations with potassium permanganate

X, 90. DISCUSSION. This valuable and powerful oxidising agent was first
introduced into titrimetric analysis by K. Margucritlc for the titration ofironflll.
In acid sohitions, the icduction can be represented by the following esjuation;

Mn 04 "-fSir TSeeiMn''

from which it foliows that the equivalent is one-fiflh of the mole, i.c. 15S.03/5.or
3 1.606. The standard potential in acid solution. /:'\has Iveen calculated lobe 1.5 1
volts, hence the permanganate ion m acid solution is a strong oxidising ucent.

Sulphuric acid is the most suitable acid, as it has no action U{K)n permanganate
in dilute solution. With hydrochloric acid, there is the likelihood of the reaction:

2MnO/ -f lOCl" Tl6ir r-:Mn=* +5C1:T8H.0

taking place, and .some permanganate may be consumed in the formation of
chlorine. Tliis reaction i.s particularly liable to occur with iron salts unless special
prccaiiiions arc adopted (see below). With a small excess of free acid, a very dilute
solution. low temperature and slow titration with constant shaking, the danger
from this cause is minimised. Tltcrc arc, however, some titrations, such as those
with ar.scnic(fll) oxide, trivalent antimony, and hydrogen i>croxide, which can be
carried out in the presence of hydrochloric acid.

In the analysis of iron ores, solution is frequently cITccted in concentrated
hydrochloric acid; the iron(Iil) is reduced and the iron(Tn is then determined in
the resultant solution. To do this, it is best to add about 25 enr' of Zimmcmanii
and Rcmhardi's soluiioii (this is sometimes termed prevmirc which is

preparc^d by dissolving 50 gcry.stalljscd niangancse(Il) sulphate MnSOi.dlKOin
250 cm’ water, adding a cooled mixture of loo cm"’ concentrated sulphuric acid
and 300 cm"' water, followed by 100 cm’ .svrupv phosphoric acid. The
mangancse(n) sulphate lovver.s the reduction potential of the Mn 04 -Mn)!!)
couple (compare Sections II, 23 -24) and thereby makes it a weaker oxidising
agent; the tendency of the permanganate ion to o.xidise chloride ion is thus
reduced. It has been stated that a further function of the inangane.se(II) sulphate is
to supply an adequate concentration of Mir ' ions to react with any local excess
of permanganate ion. Mndll) is probably formed in the reduction of
permanganate ion to manganc-sedl): the Mn(il), and also the phosphoric acid,
exert a depressant effect upon the potential of the Mn(lll)-Mn(n) couple, so that

348

TITRIMETRIC ANALYSIS X,90

Mn(III) is reduced by Fe^ ion rather than by chloride ion. The phosphoric acid
combines with the yellow Fe^'*' ion to form the'complex ion [Fe(HP 04 )]'*', thus
rendering the end-point more clearly visible. The phosphoric acid lowers the
reduction potential of the Fe(III)-Fe(II) system by complexation, and thus tends
to increase the reducing power of the Fe^"*" ion. Under these conditions
permanganate ion oxidises iron(II) rapidly and reacts only slowly with chloride
ion. ' ■

For the titration of colourless or slightly coloured solutions, the use of an
indicator is unnecessary, since as little as 0.01 cm^ of O.OliV-potassium
permanganate imparts a pale-pink colour to 100 cm^ of water. The intensity of
the colour in dilute solutions may be enhanced, if desired, by the addition of a
redox indicator (such as sodium diphenylamine sulphohate, JV-phenylanthranilic
acid, or ferroin) just before the end-point of the reaction; this is usually not
required, but is advantageous if more dilute solutions of permanganate are used.

Potassium permanganate also finds some application in strongly alkaline
solutions. Here two consecutive partial reactions take place:

(i) the relatively rapid reaction:

MnO ^ ~ +e ^MnO^^ “
and (ii) the relatively slow reaction :

MnO^^ - -h 2 H 2 O -h 2e ^ Mn02 -t- 40H “

The standard potential of reaction (i) is 0.56 volt and of reaction (ii) 0.60 volt.
By suitably controlling the experimental conditions (e.g., by the addition of
barium ions, which form the sparingly soluble barium manganate as a fine,
granular precipitate), reaction (i) occurs almost exclusively; the equivalent is then
1 mole. In moderately alkaline solutions permanganate is reduced quantitatively
to manganese dioxide. The half-cell reaction is :

Mn04-+2H20 + 3e^Mn02+40H''

and the standard potential is 0.59 volt.

Potassium permanganate is not a primary standard. It is difficult to obtain the
substance perfectly pure and completely free from manganese dioxide. Moreover,
ordinary distilled water is likely to contain reducing substances (traces of organic
matter, etc.) which will react with the potassium permanganate to form
manganese dioxide. The presence of the latter is very objectionable because it
catalyses the auto-decomposition of the permanganate solution on standing. The
decomposition:

4Mn04-+2H20 = 4Mn02 -1-302 +40H-

is catalysed by solid manganese dioxide. Permanganate is inherently unstable in
the presence of manganese(II) ions :

2 Mn 04 +3Mn^+-t-2H20 = 5Mn02-h4H + ;

this reaction is slow in acid solution, but is very rapid in neutral solution. For
nese reasons, potassium permanganate solution is rarely made up by dissolving
weighed amounts of the highly purified (e.g., A.R.) solid in water; it is more usual
0 heat a freshly prepared solution to boiling and keep it on the steam bath for an
our or so, and then filter the solution through a non-reducing filtering medium,
such as purified glass wool or a sintered glass filtering crucible (porosity No. 4).

349

X,90 quantitative INORGANIC ANALYSIS

Allcrnativcly, the solution may be allowed to stand for 2-3 days at room
temperature before filtration. The glass-stoppered bottle or flask should be
carefully freed from grease and prior deposits of manganese dioxide; this may be
done by rinsing with dichromatc-sulphuric acid cleaning mixture and then
thoroughly with distilled vvalcr. Acidic and alkaline solutions arc less stable than
neutral ones. Solutions of permanganate should be protected from unnecessary
exposure to light; a dark -coloured bottle is recommended. Diffuse daylight
causes no appreciable decomposition, but bright sunlight slowly decomposes
even pure solutions.

Potassium permanganate solutions may be standardised using arscnicffll)
o.xidc or sodium oxalate as primary standards; secondary standards include
metallic iron, and iron(ll) cthylencdiammonium sulphate (or cthylcnediamine
ironfll) sulphate). reSOi.C-lIjfNHjl^^SOi.dHjO,

Of these substances sodium oxalate was formerly regarded ns the most
trustworthy, since it is readily obtained pure and anhydrous, and the ordinary
A.R. substance has a purity of at least 99.9 per cent. The experimental procedure
hitherto employed was due to R. S. McIIridc. A solution of the oxalate, acidified
with dilute .sulphuric acid and warmed to SO-90 ‘C. was titratcrl with the
permanganate solution slowly (10- 15 cm’ per minute) and with constant stirring
until the first pennanent faint pink colour was obtained; the temperature near the
end-point was not allowed to fall below 60 'C. R. M, Fowler and H. A. Bright
have, however, shown that with McBride’s procedure the rcsult.s may be 0.1-0.45
per cent high; the tiire depends upon the acidity, the temperature, the rate of
addition of the ivermanganaie solution, and upon the speed of stirring. These
authors recommend a more rapid addition of 90-95 percent of the permanganate
solution (about 25-35 cm^ per minute) to a solution of sodium oxalate in Af-
sulphuric .acid at 25-30 C, the solution is then warmed to 55-60 'C and the
titration completed, the last 0.5-l-cnr' portion being added dropwisc. The
method is accurate to 0,06 per cent. Full experimental details arc given in
Procedure B below.

2Na’ +Cj 04 '' +2\\' =iH,Ci 04 ->- 2 Na’

2Mn04'+5HiCj04 + 6H* =r 2Mn'' -MOCO-eSMiO

It should be mentioned that if oxalate is to be determined it is often not
convenient to use the room-temperature technique for unknown amounts of
oxalate. The permanganate solution may then be standardised against sodium
oxalate at about 80 '’C u.sing the same procedure in the standardisation as in the
analysis.

The procedure of H. A. Bright, which utilisc's arscnic(]ll) oxide as a primary
standard and pouussium iodide or potassium iodateas a catalyst for the reaction,
is more convenient in practice and is a trustworthy" method for the
standardisation of permanganate solutions. A.R. arscnic<ni)*oxide has a purity of
at least 99.8 per cent, and the results by this method agree to within 1 part in 3000
with the sodium oxalate procedure of Fowler and Bright. Full experimental
details arc given in Procedure A (Section X, 92).

Asj03+40H- = 2 HAs 03 ^- +H3O
5H3As03 + 2Mn04“+6H^ = SIljAsO^q ^Mn^’ -(- 3 H 3 O
Potassium iodide, if specially purified, may be used as a primary standard. For

350

TITRIMETRIC ANALYSIS X, 91/92

many practical purposes, the dry A.R. reagent is sufficiently pure. The
potentiometric method (see Chapter XIV) should be employed : a bright platinum
indicator electrode and a saturated calomel electrode are required. The
concentration of the sulphuric acid should be about 0.4M.

10r. + 2 MnO 4 “ + 16H+ = 5l2 + 2 Mn 2 + + 8 H 2 O

Iron wire of 99.9 per cent purity is available commercially and the A.R. reagent
is a suitable standard, particularly if the potassium permanganate solution is
subsequently to be employed in the determination of iron. If the wire exhibits any
sign of rust, it should be drawn between two pieces of fine emery cloth, and then
wiped with a clean, dry cloth before use. The reaction which occurs is :

Mn 04 - + 5Fe^++8H+ =Mn2++5Fe3++4H20

Ethylenediammonium iron(II) sulphate, FeS04,C2H4(NH3)2S04,4H20 is
relatively stable and has a high molecular weight (382.16). The preparation is as
follows.

To 10.0 g of a 99 per cent solution of ethylenediamine, add 60 cm^ 6i\r-sulphuric
acid and 46.3 g A.R. iron(II) sulphate heptahydrate. Dilute to 300 cm^ with
distilled water, and to the resulting solution introduce 300 cm^ of ethanol slowly
and with constant stirring. Filter through a sintered glass funnel, wash the
precipitate with 50 per cent ethanol, and redissolve it in slightly acidulated water.
Add two-thirds the volume of ethanol. Filter again as before, and wash the solid
successively with 65 per cent ethanol and 95 per cent ethanol. Dry in the air or at
50 °C for about 12 hours. The yield is about 50 g.

X,91. PREPARATION OF O.IN-POTASSIUM PERMANGANATE.
Weigh out about 3.2-3.25 g A.R. potassium permanganate on a watch glass,
transfer it to a 1500-cm^ beaker, add 1 litre water, cover the beaker with a clock
glass, heat the solution to boiling, boil gently for 15-30 minutes and allow the
solution to cool to the laboratory temperature. Filter the solution through a
funnel containing a plug of purified glass wool, or through a Gooch crucible
provided with a pad of purified asbestos, or, most simply, through a sintered glass
or porcelain filtering crucible or funnel. Collect the filtrate in a vessel which has
been cleaned with chromic acid mixture and then thoroughly washed with
distilled water. The filtered solution should be stored in a clean, ^ass-stoppered
bottle, and kept in the dark or in diffuse light except when in use: alternatively, it
may be kept in a bottle of dark-brown-coloured glass.

X,92. STANDARDISATION OF PERMANGANATE SOLUTIONS.
Procedure A. With arsenicflll) oxide. Dry some A.R. arsenic(III) oxide at
105-110 “C for 1-2 hours, cover the container, and allow to cool in a desic-
cator. Accurately weigh approximately 0.25 g of the dry oxide, and transfer it
to a 400-cm^ beaker. Add 10 cm^ of a cool solution of sodium hydroxide,
prepared from 20 g sodium hydroxide and 100 cm^ water (1). Allow to stand for
8-10 minutes, stirring occasionally. When solution is complete, add 100 cm^
water, 10 cm^ pure concentrated hydrochloric acid, and 1 drop 0.0025M-
potassium iodide or potassium iodate (2). Add the permanganate solution from a
urette until a faint pink colour persists for 30 seconds. Add the last 1-1.5 cm^
ropwise, allowing each drop to become decolorised before the next drop is
introduced. For the most accurate work it is necessary to determine the volume of

351

X, 90 QUANTITATIVE INORGANIC ANALYSIS

Alternatively, the solution may be allowed to stand for 2-3 days al room
temperature before filtration. The gla.ss-,stoppcrcd bottle or flask should be
carefully freed from grease and prior deposits of manganese dio.xide: this may be
done by rinsing with dichromate-sulphuric acid cleaning mixture and then
thoroughly with distilled water. Acidic and alkaline solutions arc less stable than
neutral ones. Solution.s of permanganate should be protected from unnecessary
exposure to light: a dark -coloured bottle is recommended. Diffuse daylight
causes no appreciable decomposition, but bright sunlight slowly decomposes
even pure solutions.

Potassium permanganate solutions may be standardised using arsenicflll)
oxide or sodium o.xalatc as primary standards: secondary standards include
metallic iron, and ironfll) cthylcnediammoniiim sulphate for cthyienetiiaminc
iron(II) sulphate), FcSO.t.CjH.sfNHjljSOa.'in-O.

Of these substances sodium oxalate was formerly regarded as the most
trustworthy, since it is readily obtained pure and anhydrous, and the ordinary
A.R. substance has a purity of al least 99.9 per cent. The experimental procedure
hitherto employed was due to R. S. McBride. A solution of the oxalate, acidified
with dilute sulphuric add and warmed to kO-90 ‘C, was titrated with the
permanganate solution slowly (10-1 5 cm-' per minute) and with con.stant stirring
until the first permanent faint pink colour was obtained; the temperature near the
end-point was not allowed to fall below f>0 C. R. M. Fowler and H. A. Bright
have, however, sliown that with McBride’s procedure the results may be 0.1-0.45
per cent high; the litre depends upon the acidity, the temperature, the rate of
addition of the permanganate solution, and upon the speed of stirring. These
authors recommend o uiore ruph! addition of 90 • 95 per cent of the permanganate
solution (about 25- .35 enr’ per minute) to a solution of sodium oxalate in Af-
sulphuric acid at 25-30 C, the solution is then warmed to 55-60 'C and the
titration completed, the last 0.5-1-cm'' portion being added dropwisc. Die
method is accurate to 0.06 per cent. I'lill experimental details arc given in
Procedure B below.

2Na’ -t-CjOr" ■*■21!’^ ?i:U;C,0..*f2Na’

2MnO,, + 5 H;CX') 4 + 6 H‘ ^2Mn==’ -f!0CO^+SH,O

It should be mentioned that if oxalate is to be determined it is often not
convenient to use the room-tcmperatiirc technique for unknown amounts of
oxalate. Die permanganate solution may then be standardised against sodium
oxalate at about 80 ‘■'C using the same procedure in the standardisation as in the
analysis.

Die procedure of H. A. Bright, which utilises ar.scnicflll) oxide as a primary
•Standard and poia.ssium iodide or potassium iodatc as a catalyst for the reaction,
is more convenient in practice and is a trustworthy' method for the
standardisation of permanganate solutions. A.R.arscnic<lIl)oxidchasapurityof
at least 99.8 percent, and the results by this method agree to within 1 part in 3000
with the sodium oxalate procedure of r'owier and Bright. Full experimental
details arc given in Procedure A (Section X, 92).

As,Oj -t-40H ' = 2 HASO 3 - - + HjO
SHjAsOj -f 2 Mn 04 ' + 6 H = SH^AsO^ + 2Mn- -t-3HjO

Potassium iodide, if specially purified, may be used as a primary standard. For

350

X. 92 QUANTITATIVE INORGANIC ANALYSIS

permanganate solution required to duplicate the pink colour at the end-point.
This is clone by adding permanganate solution to a .solution containing the same
amounts of alkali, aedd, and eataly.st us were used in the test. Tlic correction
should not be more than 0.03 cm-*. Repeat the determination with two other
similar quantities of oxide. Calculate the normality of the {Wtassium
permanganate solution. Duplicate determinations should agree within 0.1 per
cent.

Notes. 1. For ckm'tuary suukuts, it is sufTicicnl to weigh out accurately
about 1.25 g of A.R. arscnic(l 11) oxide, dissolve this in 50 cm’ of a cool 20 per cent
solution of sodinrn hydroxide, and make up to 250 cm’ in a graduaii^ flask.
Shake well. Measure 25.0 cm-' of this solution by means of a burcttcand not with a
pipette (caution— the solution K highly poisonous) into a 250'350-an’ conical
flask, add 100 env’ water, 10 enr' pure concentrated hydrochloric acid, 1 drop
pota.ssium iodide solution, and titrate with tlic permanganate .solution to the first
permanent pink colour ns detailed above. Repeat with nvo other 25-cm’ portions
of the solution. Successive titrations should agree within O.I cm’.

2. 0.0025A/-potassium iotlide — 0.41 g KI dm' ', 0.0025.^/-potassium
iodate = 0.54gK10j dm' ’.

Calculation. It is evident from the equation given in Section 90 and
al.so from the equation:

AsjO, 4 20 = AsjO,

that the equivalent of ar.scnictll!) oxide is one quarter of a mole, 197.84/4 or
49.460 g. One cm’ of a normal solution contains the milli-cquiva!ent,or0.04946g.
If the weight of arsenictlll) oxide be divided by the nuntber of cm’ of potassium
permanganate solution to which it is equivalent as found by titration, we have the
weight of primary standard equivalent to I cm ’ of the permanganate solution. If
this last value be divided by the milli-cquivalent of ar.senic(in) oxide, the
normality of tb.e permanganate solution is obtained.

Procedure R. With sodium oxal.ate. Dry some A.R. sodium oxalate at
105-1 10 “C for 2 hours, and allow it to cool in a covered vessel in a de.siccator.
Weigh out accurately from a weighing bottle about 0.3 g of the dry sodium
oxalate into a 600-cm-' bc.aker. add 240 cm’ of recently prepared distilled water,
and 12.5 cm-' of concentrated sulpliuricacid iniurion) or 250 cm’ or2A’-sulphuric
acid. Cool to 25-30 C and stir until the oxalate has di.ssoivcd (1), Add 90-95 per
cent of the required quantity of permanganate solution from a burette at a rate of
25-35 cm’ per minute while stirring .slowly (2), Heat to 55-60 C fuse a
thermometer a.s stirring rod), and complete the titration by adding permanganate
solution until a faint pink colour persists for 30 seconds. Add tlm last 0.5-1 cm’
dropwi.se, with particular care to allow each drop to become decolorised before
the next is introduced. For the most exact work, it i.s ncce.ssary to determine the
excess of permanganate solution required to impart a pink colour to the solution.
This is done by matching the colour produced by adding permanganate solution
to the .same volume ofboilcd and cooled diluted stilpburic acid at 55-60 “C. This
correction iisiinlly amounts to 0.03-0.05 cm’. Repeat the determination with two
other similar quantities of sodium oxalate.

Noic.s. 1, For clcmeinary students, it is suflicicnt to weigh out accurately
about 1.7 g of A.R. sodium oxalate, transfer it to a 2S0-cnv’ graduated flask, and
make up to the mark. Shake well. Use 2501-0’ of this solution per titration and add
150 cm of ca. A/-sulphuric acid. Carry out the titration rapidly at the ordinary

352

TITRIMETRIC ANALYSIS X, 92

temperature until the first pink colour appears throughout the solution, and
allow to stand until the solution is colourless. Warm the solution to 50-60 °C and
continue the titration to a permanent faint pink colour. It must be remembered
that oxalate solutions attack glass, so that the solution should not be stored more
than a few days. , * _

2. An approximate value of the volume of permanganate solution required can
be computed from the weight of sodium oxalate employed. In the first titration
about 75 per cent of this volume is added^ and the determination is completed at
55-60 “C. Thereafter, about 90-95 per cent of the volume of permanganate
solution is added at the laboratory temperature.

Calculation. This is similar to that described under Procedure A. The
equivalent of sodium oxalate is j mole or 67.00 g.

Procedure C. With metallic iroa Use A.R. iron wire of 99.9 per cent assay
value. Insert a well-fitting rubber stopper provided with a bent delivery tube into
a 350-cm^ conical flask and clamp the flask in a retort stand in an inclined

position, the tube being so bent as to dip into a
small beaker containing saturated sodium
hydrogencarbonate solution or 20 per cent
potassium hydrogencarbonate solution (pre-
pared from the A.R. solids) (Fig. X, 17). Place
100 cm^ 3A-sulphuric acid (from 92 cm^ water
and 8 cm^ concentrated sulphuric acid) in the
flask, and add 0.5-1 g A.R. sodium hydro-
gencarbonate in two portions; the carbon
dioxide produced will drive out the air.
Meanwhile, weigh out accurately about 0.15 g of
iron wire, place it quickly into the flask, replace
the stopper and bent tube, and warm gently until
the iron has dissolved completely. Cool the flask
rapidly under a stream of cold water,* and then
run the permanganate solution cautiously from a
burette, with constant shaking, until the faint pink colour is permanent. The
addition of about 5 cm^ of pure syrupy phosphoric acid facilitates the detection
of the end-point. Repeat the determination with two other samples of the iron
Wire.

The reaction is:

%X,17

Mn04“ + 5Fe^+-f8H+ = Mn^+ -f-5Fe^’" + 4 H 2 O
1 equivalent Mn 04 “ = 1 mole Fe

. Procedure D. With ethylenediammonium iron(II) sulphate (see Section X,
'• W^igh out accurately from a weighing bottle about 1.3-1. 5 g of the salt into
fh conical flask, add 60 cm^ M-sulphuric acid, and swirl the contents of

e flask until the solid has dissolved. Titrate inunediately with permanganate
so ution to the first permanent faint pink tinge. (The addition of about 5 cm^
sympy phosphoric acid sp. gr. 1.75, facilitates the detection of the end-point),
rovided that it is stored with due regard to the precautions referred to in

solution is automatically drawn in until the pressure of
carbon dioxide inside the flask is equal to the atmospheric pressure.

353

X. 93/94 QUANTITATIVE INORGANIC ANALYSIS

Section X, 90, tlic standardised permanganate solution will keep for a long time,
hut it is advisable lo rc-standardisc the solution frequently to confirm that no
decomposition has set in.

X, 93. DETER.MINATION OF IRON(II). The detailed c.xpcrimcntal
method has already been given under Proceihirca C and I) of Section X, 92; the
.solution is acidified with dilute .sulphuric acid. If chloride ion is present high
results arc obtained, because the reaction between ironfll) and permanganate
induces the oxidation of hydrochloric acid. The chloride ion Ls rendered almost
harmless by the addition of a nianganescfll) .salt, preferably in the form of the so-
called Zinunemtann -Reinhardt or preventive solution (Section X. 90), and by
slow titration.

The test solution should be approximately O.IAf with respect to iron(ll|, and
should contain about lO per cent (by volume) of dilute sulphuric add to reduce
the tendency for atmospheric o.xidaiion of the iron solution. Pipette 25 cm- ofthc
solution into a 25f)-env' conical flask, add 25 cm-’ sulphuric acid (0.5.\f). and
titrate with the standard (0,1 A’) potassium permanganate solution unlil a faint
permanent pink coloration is produced.

A solution containing iron in the Irhaicnt condition may also l>e analysed by
titration with standard pot.issium permanganate after reduction of the iron to the
divalent condition; this may he done, preferably by use of a Jones rcductor.or by
one of the other methods discusset! in Sections X, 142- fi.

Tlic iron content of an iron ore may be similarly determined by dissolving a
known weight of the ore in dilute ludrochloric acid and making the resulting
solution up to the mark in a graduated flask. An aliquot portion of the solution is
subjected to a suitable rerluciion procedure and is then titrated with standard
permanganate. solution in the presence of Zimmennann-Rcinhardt solution.

X, 94. DCTERMIN.ATJON OK CAECIU.M. The calcium is precipitated as
oxalate, tlie washed precipitate is dissolved in dilute sulphuric acid, and the oxalic
acid liberated is titrated with standard permanganate solution. Predpilation of
the calcium oxidate is best carried out by homogeneous precipitation using the
urea hydrolysis method, in which acid is added to the solution to produce a pH of
about 1.0; this is followed by ammonium o.xalntc and urea. Upon boiling the
solution, the urea gradually undergoes hydrolysis and the pH rises to the point of
calcium oxalate precipitation; this may take 10-15 minutes. The crystals
precipitated from the hot solution are relatively large, and may be filtered off
shortly after formation; this climinate.s the digestion' period which is otherwise
required. Tlie solution must remain clear until boiling iscommenced to hydrolyse
the urea.

When .suipiiale is present, both thi.s and the normal method of precipitation
yield high rcsiilt.s in a single precipitation. With a double precipitation the error
by the urea method is considerably sniallcr and anv macnesium present is almost
completely eliminated.

Procedure. Weigh out accurately 0.15 0.20 g calcium carbonate,
preferably of A.R. grade, into a 400-cm^ beaker. Add 20 enr’ water and cover the
beaker with a clock glass. Introduce 10 cm* dilute liydrochloric acid (T. 1) and
warm, if necessary, until the solid has dissolved. Dilute’ to 200 cm*, and add a few
drops of methyl red indicator; suflicient acid must be pre.sent in the solution to
prevent the precipitation of calcium oxalate when ammonium oxalate solution is

354

TITRIMETRIC ANALYSIS X,95

added. Now introduce 15 cm^ saturated ammonium oxalate solution and 15 g
urea. Boil the solution gently until the methyl red changes colour to yellow (at pH
5). Filter through a coarse filter paper, or, with suction, on a small filter paper
supported in a Gooch crucible. Wash the precipitate with small volumes of cold
water until free from chloride. Transfer the filter paper and precipitate (or the
Gooch crucible and precipitate) to the original beaker, dissolve the precipitate in
hot dilute sulphuric acid, and titrate immediately with standard O.lN-potassium
permanganate solution : follow Procedure B, Section X, 92.

X, 95. ANALYSIS OF HYDROGEN PEROXIDE. Hydrogen peroxide is
usually encountered in the form of an aqueous solution containing about, 6 per
cent, 12 per cent or 30 per cent hydrogen peroxide, and frequently referred to as
‘20 volume’, ‘40 volume’, and ‘100 volume’ hydrogen peroxide respectively; this
terminology is based upon the volume of oxygen liberated when the solution is
decomposed by boiling. Thus 1 cm^ of ‘100 volume’ hydrogen peroxide will yield
100 cm^ of oxygen measured at s.t.p.

The following reaction occurs when potassium permanganate solution is
added to hydrogen peroxide solution acidified with dilute sulphuric acid :

2Mn04- + 5 H 202 + 6H+ = 2Mn2+ + 5O2 + 8H2O

This forms the basis of the method of analysis given below.

It is good practice to use a fairly high concentration of acid and a reasonably
low rate of addition in order to reduce the danger of forming manganese dioxide,
which is an active catalyst for the decomposition of hydrogen peroxide. For
slightly coloured solutions or for titrations with dilute permanganate, the use of
ferroin as indicator is recommended. Organic substances may interfere. A fading
end-point indicates the presence of organic matter or other reducing agents, in
which case the iodimetric method is better (Section X, 124).

Procedure. Transfer 25.0 cm^ of the ‘20-volume’. solution by means of a
burette to a 500-cm^ graduated flask, and dilute with water to the mark. Shake
thoroughly. Transfer 25.0 cm^ of this solution to a conical flask, dilute, with 200
cm^ water, add 20 cm^ dilute sulphuric acid ( 1 : 5), and titrate with standard 0. 1 N-
potassium permanganate to the first permanent, faint pink colour. -Repeat, the
titration; two consecutive determinations should agree within 0.1 cm^ ’ . ^

Calculate: (i) the weight of hydrogen peroxide per dm^ of, the original solution
and (ii) the ‘volume strength’, i.e., the number of cm^ of oxygen at s.t.p. that can be
obtained from 1 cm^ of the original solution.

A metallic peroxide, such as sodium peroxide, can be analysed in s imil ar
manner, provided that care is taken to avoid loss of oxygen during the dissolution
of the peroxide. This may be done by working in a medium containing boric acid
which is converted to the relatively stable ‘perboric acid’ upon the addition of the
peroxide.

Procedure. To 100 cm^ of distilled water, add 5 cm^ concentrated sulphuric
acid, cool and then add 5 g pure boric acid; when this has dissolved cool the
mixture in ice. Transfer gradually from a weighing bottle about 0.5 g (accurately
weighed), of the sodium peroxide sample to the well-stirred, ice-cold solution.
When the addition is complete, transfer the solution to a 250-cm^ graduated
flask, make up to the mark, and then titrate 50-cm^ portions of the solution with
standard (O.IN) permanganate solution. ■ . :

355

X, 96/97 QUANTiTATiVn INORGANIC ANALYSIS

X, 96. DETERMINATION OF MANGANISE DIOXIDE IN PVROLU-
SITE. Discti‘:xin)i. Manganese dioxide occurs in nature as the mineral
pyrolusilc. I'or many purposes, a knowledge of the percentage of MnOj is
required. Tin’s may be determined by treatment with an excess of an acidified
solution of a reducing agent, such as .<odium oxalate, or arscnicflll) oxide.

Mn0j-fHaC,04 + 2H’ - a-2COi+2H,0

2MnOj+2n.,AsO,-i ‘Ur --2Mn‘" + 2 n,,As 04 + 2H,0

The excess of reducing agent is determined by titration with .standard
permanganate solution. Arsenic! I If) oxide is somewhat more trustworthy in this
determination than i.s .sodium oxalate, because oxalic acid decomposes vciy
slowly at high temperatures into carbon monoxide and carbon dioxide, the
decomposition being catalysed by manganesefll) salts; the extent of
decomposition under ordinary circumstances is. however, very small. Both
procedurc.s will bcdescrilvd

Procedure A (arsenicdll) oxide mctliodl. Dry the finely powdered sample
of pyrolusilc at 120'C loconstant weight. Weigh out accurately from a weighing
bottle about 0.2 g of the sample into a 250-cm^ conicil flask, add 50 cm^ of
standard O.lA'-arscnictlll) oxide (ScciuMt X, 92, Note 1) and 10 cm- of
concentrated sulphuric acid. Place a .short funnel in the mouth of the fla.sk'. and
boil until the pyrolusilc has decomposed completely; no brown or black particles
should then be present. Cool the solution, add 1 <frop of 0,0025Af-potassium
iodide solution, and titrate the excess arscnictlll) oxide with standard O.IX-
potassium permanganate. Repeal the determination with two other samples of
the solid.

Calculate the percentage of MnOj in the pyrolusilc from the amount of
arscnicflll) oxide consumed in the reaction.

Procedure B (sodium oxalate nu'tliodl. Weigh out accurately about 0.2 g of
the finely powdered, dry pyrolu.sitc into a conical flask, add 50 cm^ of standard
O.lA'-sodium oxalate (Si-elion X, 92, /). Note 1). add 50 cm^ of2iV/-.sulphuricacid
(<Vi. 10 per cent), and place a short funnel in the mouth of the flask. Boil the
mixture gently until no black particles remain. Allow to cool, and titrate the
excess of oxalate with standard O.liV-potnssium {scrmanganaie as detailed in
Section X, 92, Procedure B. Repeat the determination with two other samples of
similar weight.

Calculate the amount of sodium oxalate consumed in the reaction, and from
this the percentage of MnOj in the pyrolusilc.

X, 97. DETERMINATION OF NTTRITFS. Discussmt. Nitrites react in
warm acid solution {ca. 40 ' C) svith permanganate .solution in accordance with
the equation;

2Mn04 ' + 5NO, ' + 6H ' 2Mn* •* + 5NO3 ' 4- 3HiO

If a solution of a nitrite is titrated in the ordinary way with potassium
permanganate, poor results arc obtained, because the nitrite solution has first to
be acidiOed with dilute sulphuric acid. Nitrous acid is liberated, which being
volatile and unstable, is partially lost. If, however, a measured volume of standard
potassium permanganate solution, acidified with dilute sulphuric acid, is treated
with the nitrite solution, added from a burette, until (lie permanganate is just

356

TITRIMETRIC ANALYSIS X, 98

decolorised, results accurate to 0.5-1 per cent may be obtained. This is due to the
fact that nitrous acid does not react instantaneously with the permanganate. This
method may be used to deterniine the purity of commercial potassium nitrite.

Procedure. Weigh out accurately about 1.1 g of commercial potassium
nitrite, dissolve it in cold water, and dilute to 250 cm^ in a graduated flask. Shake
well. Measure out 25.0 cm^ of standard O.liV-potassium permanganate into a
500-cm^ flask, add 300 cm^ of 0.75iV-sulphuric acid, and heat to 40 °C. Place the
nitrite solution in the burette, and add it slowly and with constant stirring until
the permanganate solution is just decolorised. Better results are obtained by
allowing the tip of the burette to dip under the surface of the diluted
permanganate solution. Towards the end the reaction is sluggish, so that the
nitrite solution must be added very slowly.

More accurate results may be secured by adding the nitrite to an acidified
solution in which permanganate is present in excess (the tip of the pipette
containing the nitrite solution should be below the surface of the liquid during the
addition), and back-titrating the potassium permanganate with a solution of
ammonium iron(II) sulphate which has recently been compared with the
permanganate solution.

X,98. DETERMINATION OF PERSULPHATES. Discussion. Alkali
persulphates (peroxydisulphates) can readily be evaluated by adding to their
solutions a known excess of an acidified iron(II) salt solution, and determining
the excess of iron(II) by titration with standard potassium permanganate
solution.

S^Og^- +2Fe=^+ +2H+ = 2 Fe 3 + -1-2HSO4"

By adding phosphoric acid or hydrofluoric acid, the reduction is complete in a
few minutes at room temperature. Many organic compounds interfere.

Another procedure utilises standard oxalic acid solution. When a sulphuric
acid solution of a persulphate is treated with excess of standard oxalic acid
solution in the presence of a little silver sulphate as catalyst, the following reaction
occurs:

H2S2O8 + H2C2O4 = 2H2SO4+2CO2

The excess of oxalic acid is titrated with standard potassium permanganate
solution.

Procedure A. Prepare an approximately O.lN-solution of ammonium
iron(II) sulphate by dissolving about 9.8 g of the A.R. solid in 200 cm^ of sulphuric
acid (0.5M) in a 250-cm^ graduated flask, and then making up to the mark with
freshly boiled and cooled distilled wafer. Standardise the solution by titrating
25-cm^ portions with standard potassium permanganate solution (0.1 N) after the
addition of 25 cm^ sulphuric acid (0.5M).

Weigh out accurately about 0.3 g potassium persulphate into a conical flask
and dissolve it in 50 cm^ of water. Add 5 cm^ syrupy phosphoric acid or 2.5 cm^
35-40 per cent hydrofluoric acid, 10 cm^ 5iV-sulphuric acid, and 50.0 cm^ of the
ca. O.lN-iron(II) solution. After 5 minutes, titrate the excess of Fe^+ ion with
standard O.lN-potassium permanganate.

From the difference between the volume of O.lN-permanganate required to
oxidise 50 cm^ of the iron(II) solution and that required to oxidise the iron(II) salt

357

X, 99 OUANTITATIVB INORGANIC ANALYSIS

remaining after the addition of the persulphate, calculate the percentage purity of
the sample,

Proccihire B. Prepare an appro.ximntely O.IN solution oxalic acid by
dissolving about 1.6 g of the A.R. material and making up to 250 cm^ in a
graduated flask. Standardise the solution with standard (O.JjV) potassium
permanganate solution using ilie procedure described in Section X, 92, B.

Wcigli out accurately 0.3 -0.4 g potassium persulphate into a 350-cm^ conical
flask, add 50 cm’ 0. 1 jV-o.xaiic acid, followed by 0.2 g of silver sulphate dissolved in
20 env’ 10 per c-enl sulphuric acid. 1 feat the mixture in a water bath until no more
carbon dioxide is evolved (15-20 minutes), dilute the solution to about 100cm’
with water at about 40 C, and titrate the cxccns of o.x;ilic add with standard
0. 1 A’-potassium permanganate.

X, 99. DKTERMINA'l tON OE MANGANESE IN STEEL A. Bismuthatc
method, Disnissmn. Tlic steel is dissolved in nitric acid and the resulting
cooled solution is treated with .sodium bismuthuic when permanganic add is
formed :

2Mn’' +5NaniO,.t 1411’ - 2MnO*"-i 5Hi” -t-7H;0 + 5Na’

Frxeess bi.smuthatc is removed by filtration, a measured volume fcxccss) of a
siand;irdiscd ammonium ironitl) siilpliaie solution i.s added to reduce the
|>crnianganic ticid, and the excess ironfl!) is then determined by titration with
standard potassium fHTrnang.inate. The solution should be free from cobalt,
chromium and chloride, and since many steels contain chromium in addition to
manganc'-e, Proa'durr B is of more general application.

H. Persulphate' arsenite method. Di.seussion. Manganese salts are
oxidised to permang.inic ticid by persulphate in tltc presence of silver nitrate
solution as catalyst :

2Mn’ ’ -f 5S,.0,'’ • -t- SILO 2,MnO., ' f lOSO.,- " + 1 6H ’

If the oxidation with persulphate is ctirried out in the presence of pho.sphoricarid,
it is possible to oxidise as nuidi as 50 mg of mangtinc.sc to pcrmiinganic acid
without the separation of oxides of manganese. No .stuisfactory method is known
for removing the excess of pcrsulplialc. boiling will destroy it. but some
permanganic acid will be decomposed at the same time. Use is mtidc of the fact
that an arsenite solution reacts rapidly with permanganic acid in thccold. but no
reaction occurs with tlic pcrsulpiiate. A little diloride is added to precipitate the
silver catalyst and thus prevent the re-oxidation of the manpane.sc(ll)-‘''''h formed
by reduction. The reduction of the jicrnianganic acid by the arsenite docs not
proceed compleidy to bivalent manganese, and it is therefore advisable to
.standardise the arsenite solution against a steel of known manganese content.

Chromium is oxidised to chromate, but the vcliow colour has little effect if the
ciiromium content does not exceed 10 mg per ioo env'. ^ ,

Procedure. Weigh out accurately about 1.0 g of the steel into a 350-cm'
conical flask and add successively 15 cm’ witcr, 3 cm’ concentrated sulphuric
acid, 4 cm’ 85 per cent phosphoric acid, and S cm’ concentrated nitric acid. Heat
until solution is complete, and boil to expel o.xides of nitrogen. Add 50cm’ water,
5 cm’ O.EM-silver nitrate solution, and 2.5 g pure ammonium persulphate
dissolved in a little water. Heat to boiling and boil briskly for 1 minute. Coo!
rapidly to 25 ' C or lower, add 75 cm’ cold water, and 5 cm’ 0 . 2 . 3 /-sodium

358

TITRIMETRIC ANALYSIS X, 100

chloride solution. Titrate immediately with 0.025Al-sodium arsenite solution (1)
to a clear yellow end-point which does not change upon the addition of more
arsenite solution.

Standardise the arsenite solution against a similar steel of known manganese
content.

Note. ' 1. Prepare the 0.025N-sodium arsenite solution by dissolving 1.230 g
A.R. arsenic(in) oxide in a solution of 10 g A.R. sodium hydroxide in 30 cm^
water, warming if necessary. Dilute to about 500 cm^, neutralise by the addition
of 29-30 cm^ IM-hydrochloric acid, then add 10 g A.R. sodium hydrogen-
carbonate, and dilute to 1 dm^ in a graduated flask.

Once the oxidation to permanganic acid has been effected, the determination
may be completed more rapidly spectrophotometrically (see Section XVIII, 23),
and an alternative procedure is to carry out a potentiometric titration in which
the solution containing Mn(II) is titrated with a standard solution of potassium
permanganate (Section XIV, 28).

Oxidations with potassium dichromate

X, 100. DISCUSSION. Potassium dichromate is not such a powerful
oxidising agent as potassium permanganate (compare reduction potentials in
Table II, 4 in Section H, 23), but it has several advantages over the latter
substance. It can be obtained pure, is stable up to its fusion point, and is therefore
an excellent primary standard. Standard solutions of exactly known strength can
be prepared by weighing out the pure dry salt and dissolving it in the proper
volume of water. Furthermore, the aqueous solutions are stable indefinitely if
adequately protected from evaporation. Potassium dichromate is used only in
acid solution, and is reduced rapidly at the ordinary temperature to a green
chromium(III) salt. It is not reduced by cold hydrochloric acid, provided the acid
concentration does not exceed 1 or 2M. Dichromate solutions are also less easily
reduced by organic matter than are those of permanganate and are also stable
towards light. Potassium dichromate is therefore of particular value in the
determination of iron in iron ores; the ore is usually dissolved in hydrochloric
acid, the iron(III) reduced to iron(II), and the solution then titrated with standard
dichromate solution:

CrjO,^- +6Fe2+ -|- 14H+ = 2Cr3+ -l-6Fe3+

In acid solution, the reduction of potassium dichromate may be represented as :

CrjO^^ - + 14H + -h 6c ^ 2Cr3 + H- 7 H 2 O

from which it follows that the equivalent is one-sixth of the mole, i.e;, 294.18/6 or
49.030 g. A O.lJV-solution therefore contains 4.9030 g dm“^.

The green colour due to the Cr^"^ ions formed by the reduction of potassium
dichromate makes it impossible to ascertain the end-point of a dichromate
titration by simple visual inspection of the solution and so a redox indicator must
be employed which gives a strong and unmistakable colour change; this
procedure has rendered obsolete the external indicator method which was
formerly widely used. Suitable indicators for use with dichromate titrations
include N-phenylanthranilic acid (0.1 per cent solution in 0.005M-NaOH) and
sodium diphenylamine sulphonate (0.2 per cent aqueous solution); the latter
must be used in presence of phosphoric acid.

359

X, 101/102/103 OUANTITATIVn INORGANIC ANALYSIS

X, 101. PREPARATION OF O.IN-POTASSIUM DICHROMATE. A.R.
potassium dichromatc Ims .a purity of not less than 99.9 per cent and is
satisfactory for most purposes.* Powder finely about 6 p of the A.R. material in a
glass or agate mortar, and heat for .30-60 minutes in an air oven at 140 -150 T.
Allow to cool in a closed vessel in a desiccator. Weigh out accurately about 4.9 g
of the dry potassium dichromatc into a weighing bottle and transfer the salt
quantitatively to a l-dm' graduated flask, using a small funnel to avoid loss.
Dissolve the'salt in the flask in water and make up to the mark; .shake well.
Alternatively, place a little over 4.9 g of potassium dichromate in ;i weighing
bottle, and weigh accurately. Empty the .salt into a I -dm-' graduated flask, and
weigh the bottle again. Llissolvc tlic salt in water, and make up to the mark.

The normality of the .solution can be calculated directly from the weight of salt
taken, but if the salt has only been w eighed out approximately, then the solution
nui.st be standardi.scd as in the following Section.

X. 102. STANDARDlS.VnON OF POTASSIUM DlCMRO.M..VrE .SOLU-
TION .\GAINST IRO.N. Use tlie method dc.scril'cd in Section X, 92. Vrocedme
C. with 0.2 g accurately weighed, of .A.R, iron wire. Titrate the cooled solution
immediately with the dichromatc .solution, using either sodium diplienyl.amtne
suiphemate or N-p!',cnylaniluaniIicacid as indicator. If the former is selected, add
6-8 drops of the indicator, folloxvcd by 5 enr' of syrupy phosphoric acid: titrate
.slowly with the dichromatc solution, stirring well, until the pure green colour
clianges to a grey-green, Tlicn add the dichromatc solution dropwise until the
first tinge of blue-violet, which remains permanent on shaking, appears. If the
latter indicator is ■>clccied. add 2CK>cm'' of .^/-sulphuric acid, then 0.5 cm^ of
the indicator; add the dichromatc .solution, with shaking until the colour
changes from grexm to violet-red.

] mole K.Cr.O, 6 moles Fc

The standardisation max also be cfTectcd with ethylcnediammonium ironlll)
sulphate as described in Section X, 92, Procfilure D.

X, 103. DErERMlN.ATlON OFIRONiIl). Tlie conditions arc very similar
to those outlined in Section X, 93 with the e.xccption that tlie presence of
moderate amounts of chloride ion have no cfTeet on the determination.

7 he test solution should be approxiniatclx 0. L\f with respect to ironfll), and
.should contain dilute sulphuric acid to rcsiuec the tendency for atmospheric
oxidation. Titrate 25.0-cm^ portions of this soliilion with the standard (O.liS’)
potassium dichromatc solution using citJicr sodium diphcnvlatnine sulphonatc
(]) or N-phcnylanthranilic acid (I I) as interna! indicator.

Use 8 drops (say 0.4 cm^) of the indicator I, add 200 cm’ of 2.5 per a'nt
sulphuric acid, followed by 5 enf’ of 85 per cent pliosphoric acid, and titrate
slowly, wliilst stirring constantly, with the standard dichromatc until the solution

If only a pure grade (a*, doiinc! from A.R.tofcommcioal s;\!s Kas.idal'tc.or if there is some doubt
as to ihc purity of the s.ilt. ihc follossin.o method of piiriCicalion should t>e used. A concentralcd
solution of the Mil m liol w.itcr is ptep-sred .nnd littered ThcctysUls sshich separate on cooling arc
fdicred on a .sintered gl.tss lilicr funnel and sitchcd dry The resultant cr\ stats arc rccrysl.itli.scd
again. Tlie purified crystals are tiicn dried at lS0-;iX) C, ground to a line oosvdcr in a glass or agate
mortar, and again dried at HO - 150 C lo constant weight.

360

TITRIMETRIC ANALYSIS X, 104

assumes a bluish-green or greyish-blue tint near the end-point. Continue the
titration, adding the dichromate solution dropwise and maintaining an interval
of a few seconds between each drop, until the addition of 1 drop causes the
formation of an intense purple or violet-blue coloration, which remains
permanent after shaking and is unaffected on further addition of the dichromate.

Use 0.5 cm^ of indicator II. Add about 200 cm^ of M-sulphuric acid and then
titrate with the O.lN-potassium dichromate until the colour changes from green
to violet-red. This titration is sharp to within 1 drop.

A solution containing iron in the trivalent condition may be analysed after
reduction of the iron to the divalent condition with a Jones redactor, or by one of
the other methods described in Sections X, 143-5.

The iron content of an iron ore may be similarly determined by weighing out
about 2 g, dissolving in dilute hydrochloric acid, and making up to the mark in a
250-cm^ graduated flask. Portions of the solution (25.0 cm^) are then subjected to
a suitable reduction procedure, and the solutions titrated with standard
potassium dichromate.

The iron ore will usually contain both iron(II) and iron(III) compounds, and
the procedure just described measures the total iron content of the ore. The
proportion of iron(II) can be determined by the following procedure.

Fit a 350-cm^ conical flask with a rubber stopper carrying two glass tubes bent
at right angles, one passing to the bottom of the flask and the other ending just
inside the stopper. Join the longer tube to a gas wash-bottle which is connected to
a source of carbon dioxide (Kipp’s apparatus or a cylinder of the compressed gas),
and attach a gas bubbler to the shorter tube; the wash bottle and the bubbler
contain distilled water. Weigh out accurately about 0.4 g of the finely powdered
ore into the flask, and then pass a stream of carbon dioxide through the flask to
displace the air; this ensures that the iron(Il) chloride formed in the subsequent
dissolution of the ore does not undergo atmospheric oxidation. Open the stopper
of the flask momentarily and introduce 30 cm^ of 1 : 1 hydrochloric acid, then
warm the flask gently and pass a slow stream of carbon dioxide until the ore has
been completely attacked ; in most cases a small white residue of silica will remain.
Allow the flask to cool whilst still maintaining the current of carbon dioxide, then
wash down the tubes and neck of the flask with a little cold, air-free distilled
water, add 200 cm^ of 2.5 per cent sulphuric acid (prepared with air-free water),
and then titrate with standard potassium dichromate solution using an internal
indicator. The iron(II) content of the ore thus determined, subtracted from the
total iron, gives the iron(III) content of the ore.

X, 104. DETERMINATION OF CHROMIUM IN A CHROMIUM(ra)
S^T. Discussiott. Chromium(III) salts are oxidised to dichromate by boiling
with excess of a persulphate solution in the presence of a little silver nitrate
(catalyst). The excess of persulphate remaining after the oxidation is complete is
destroyed by boiling the solution for a short time. The dichromate content of the
resultant solution is determined by the addition of excess of a standard iron(II)
solution and titration of the excess of the latter with standard O.lAT-potassium
dichromate.

2Cr3+ 38208^- +6HSOr +8H+

+ 2 H 2 O = 02t + 4HS04-

361

X. 105 QUANTITATIVE INORGANIC ANALYSIS

Prtycedurc. Weigh out accurately an amount of the salt which will contain
about 0.25 g ofehromium, and dissolve it in 50 cm ’ distilled water. Add 20 cm’ of
CO. O.IAf-silvcr nitrate solution, followed by 50 cm’ of a 10 per cent solution of
ammonium or potassium persulphate. Roil the liquid gently for 20 minutes. Cool,
and dilute to 250 cm’ in a graduated flask. Remove 50 cm’ of the solution with a
pipette, add 50 cm’ ofO.IA'-ammonium ironfl!) sulphate .solution (Section X. 98.
Procedure .-It, 200 cm’ of 2iV-sulp!mric acid, and 0,5 cm’ of iV-phcny!anthrani!ic
acid indicator. Titrate thec\cc.ss of the ironfll) salt with standard O.lA'-pota.ssium
dichromatc until the colourchangcs from green to violet-red.

Slandardi.se the ammonium ironflll sulphate solutioti against the O.LV-
potassium dichromate, using ;Y-phcnylanthranilic acid as indicator. Calculate
the volume of the irontll) solution which was oxidised by the dichromatc
originating from the chromium .stilt, and from this the percentage of chromium in
the sample.

Note. Lead or Iiarlum can he determined by precipitating the sparingly
soluble cliromate. dissolving the washed prccipittite in dilute sulphuric acid,
adding a known c.xcess of ammonium ironfl 1) sulphate solution, and titrating the
excess of Fc’ " ion with 0, 1 .V-poi;issium diehrom.atc in the usual way.

2PbCrO^-i 2ir - 2I'b’* -!-Cr,0.’~ -r M;0

X, U)5. DETERMINATION OF CHROMIUM !N CHROMITIC. Dheus-
iion. The highly rcfriiciory mineral chromite is brought into solution by fusion
with c.xce.ss of sodium peroxide.

2Fc(CrOj); + 7Na ,0, r. 2NaFc(); T -INaXrOj + 2Na,0

or 2Fc(CrO,); -t 70.’‘ -- 2FcO, •• -t-dCrO^’' +20’“

Upon leaching the melt with water, the sodium chromate dissolves and the iron is
precipitated as irontlll) hydroxide:

NaFcO- t 211,0 rr, NaOH -t FetOH),,

2Na, 0 + 211,0 ^ -tNaOIi

The cxccs,s of peroxide is deconijmsed by boiling the alkaline solution. ITic
precipitate ts filtered off after diluting the solution; the filtrate is acidified with
hydrochloric acid, a known solumc of excess of ammonium ironflll .sulphate
.solution is added, and the excess of ironfl I) is titrated with standard potassium
dichromatc solution.

2CrO,’ ' +2H ‘ ^Cr.Or' ' -i H,0

Cr,0-’- +6Fc’" 4. Mir 2Cr” +f,Fc" +711,0

Procedure. Weigh out accurately about 0,5 g of the very finely powdered
ore into a oO- 35-cni’ nickel, or heavy-walled porcelain, crucible, add 4 g of
sodium peroxide, and mix thoroughly by means of ;i thin class rod. Remove any
powder adhering to the rod by stirring about 1 g of .sodium psToxide with it;covcr
the mixture in the crucible with this peroxide. Place the lid on the crucible, and
gently heat the covered crucible in the fume cupboard over a .small flame until the
mass is quite liquid (about 10 minutes); keep fused for a further 10 minutes at a
dull red heat. Allow to cool, and when a solid crust has formed, add 4 g more of
the sodium peroxide, and fu.se the mixture again at a cherry-red heat for 10

362

TITRIMETRIC ANALYSIS X, 106/107

minutes. Allow the crucible to cool and place it in a 600-cni® Pyrex beaker
containing a little distilled water. Cover the beaker with a clock glass, add a little
warm water, and, after the violent action has subsided, remove the crucible and
wash it thoroughly, collecting the washings in the same beaker. Boil the liquid for
30 minutes, keeping the beaker covered (this decomposes the hydrogen peroxide),
add 250 cm^ boiling water, and allow the precipitate to settle. Filter through a
hardened 15-cm filter paper or, better, through a sintered glass filtering crucible,
and wash the residue thoroughly with boiling water until free frorh chromate.
(The residue should be completely soluble in concentrated hydrochloric acid; no
black gritty particles should remain. If this is not the case, decomposition is not
complete, and the determination must be started afresh.) Evaporate the filtrate to
about 200 cm^, cool, and add 4.5M-sulphuric acid cautiously until acid. Cool,
transfer to a 250-cm^ graduated flask, and make up to the mark with distilled
water. Shake well. Remove 50 cm^ of this solution with a pipette, add 50 cm^ of
O.lAT-ammonium iron(II) sulphate and proceed as in the previous Section.

X,106. DETERMINATION OF CHLORATE. Discussion. Chlorate ion is
reduced by warming with excess of iron(II) in the presence of a relatively high
concentration of sulphuric acid :

C 103 - + 6Fe2++6H+ =Cl-+6Fe3+ + 3H20

The excess Fe^'*' ion is determined by titration with standard dichromate
solution in the usual way.

Procedure. To obtain experience in the method, the purity of A.R.
potassium chlorate may be determined. Prepare a 0.02M-potassium chlorate
solution using the A.R. solid. Into a 250-cm^ or 350-cm^ conical flask, place 25.0
cm^ of the potassium chlorate solution, 25.0 cm^ of 0.2N-ammonium iron(II)
sulphate solution in 4N-sulphuric acid and add cautiously 12 cm^ concentrated
sulphuric acid. Heat the mixture to boiling (in order to ensure completion of the
reduction), and cool to room temperature by placing the flask in running tap
water. Add 20 cm^ l:l-phosphoric acid, followed by 0.5 cm^ sodium
diphenylaminesulphonate indicator. Titrate the excess Fe^'^ ion with standard
O.liV-potassium dichromate to a first tinge of purple coloration which remains on
stirring.

Standardise the ammonium iron(II) sulphate solution by repeating the
procedure but using 25 cm^ distilled water in place of the chlorate solution. The
difference in titres is equivalent to the amount of potassium chlorate added.

Oxidations with ceriuni(IV) sulphate solutions

X,107. GENERAL DISCUSSION. Cerium(IV) sulphate is a powerful
oxidising agent; its reduction potential in l-8N-sulphuric acid at 25 °C is 1.43
±0.05 volts. It can be used only in acid solution, best in 0.5 N or higher
concentrations: as the solution is neutralised, cerium(IV) hydroxide (hydrated
cerium(IV) oxide) or basic salts precipitate. The solution has an intense yellow
colour, and in hot solutions which are not too dilute the end-point may be
detected without an indicator; this procedure, however, necessitates the
application of a blank correction, and it is therefore preferable to add a suitable
indicator.

The advantages of cerium(IV) sulphate as a standard oxidising agent are:

363

X, 107 QUANTITATIVE INORGANIC ANALYSIS

1. Ccrium{lV) sulphate solutions arc remarkably stable over prolonged
periods. Tlicy need not be protected from liphU and may even be boiled fora short
time without appreciable change in concentration. Tlic stability orsulphuric add
solutions covers the wide range of lO^dO cm* of concentrated sulphuric acid per
lilrc.lt is evident, therefore, that an acid solution ofccrium{IV)sii!phate surpasses
a permanganate solution in stability.

2. Cerium(lV) sulphate may be employed in the determination of reducing
acents in the presence of a high concentration of hydrochloric acid (contrast
potassium permanganate. Section X,90).

3. Ccrium(IV) solutions in 0.1 iV solution are not too highly coloured to
obstruct vision when reading the meniscus in burettes and other titrimciric
apparatus.

4. In the reaction ofccriumdV) salts in acid solution with reducing ageni.s, the
simple valency change

CC** + c=iCc’"

is assumed to lake place; the equivalent sveight is therefore the mole. With
permanganate, of course, a number of reduction products arc produced
•according to the experimental conditions,

5. The ceriumdll) ion is colourle.ss (compare colourlc.ss mangancsdll) ion
from potassium permanganate, and green diromiumfllll ion from poMssium
dichromate).

(). Ccrium(lV) .sulphate is a very vcr.saiilc oxidising agent. It may be employed
in most titrations in xvhicli permanganate has Ivecn used, and also for other
determinations.

7. Ccrium(lV) sulphate solutions arc best standardised with ar.scnic(lll)o.xide
or with sodium oxalate.

Solutions of cerium(IV) sulphate in dilute sulphuric add are stable even at
boiling temperatures. I lydrochlot ic acid solutions of the salt arc unstable because
of reduction to ccrium(lll) by the acid with the simulianeou.s liberation of
chlorine;

2Cc^* +2C1 -- 2Ce*' -i Cl;

This reaction lakes place quite rapidly on boiling. :md hence hydrochloric .acid
cannot be used in oxidations which necessitate boiling with excess of cerium(iV)
sulphate in acid solution: sulphuric acid must l>e used in such oxidations.
However, direct titration with ccriiim(IV) sulphate in a dilute hydrochloric .add
medium (c.g., for iron(II) may be accurately performed at room temperature, and
in this respect cerium(lV) sulphate is superior to potassium permanganate (cf. 2
above). The presence of hydrofluoric acid is harmful, .since (iuoridc ion forms a
stable complex wiiii Cc(I\') and decolorises the yellow solution.

Formal potential nicasiirement.s show that the redox potential of the Cc(l'0“
Cc(III) sy.stcm is greatly dependent upon the nature and the concentration of the
acid present; thus the foilowinc values arc recorded for the acids named in molar
•solution .•H,.S04 1.^14 V.IINO', 1.61 V.HClOj 1.70 V. and in SA/ perchloric .acid
solution the value is 1,87 V.

It has been postulated on the basis of the forma! potential measurements that
Cc(l V) cxist.s as anionic complexes [CcfSOjlJ" ' or (CcfSOjlj]-’”. [Ce(NOj)6l*".
iind [Cc(C104)(,]' ; in consequence, solid salts such a.s ammonium ccriumflV)
sulphate 3(NH4),S04 .Cc(S 04).,2 Hj 0 and ammonium ccriumflV) nitrate

364

TITRIMETRIC ANALYSIS X, 108

2NH4N03.Ce(N03)4,4H20 have been formulated as ammonium tetrasulph-
atocerate(IV) (NH4)4[Ce(S04)4]2H20 and' ammonium hexanitratocerate(IV)
(NH4)2[Ce(N03)6]4H20 respectively. For convenience, the term cerium(IV)
sulphate will be retained.

Solutions of cerium(IV) sulphate may be prepared by dissolving cerium(IV)
sulphate or the more soluble ammonium cerium(IV) sulphate in dilute (I-IN)
sulphuric acid. Ammonium cerium(IV) nitrate may be purchased of A.R. quality,
and a solution of this in M-sulphuric acid may be used for many of the purposes
for which cerium(IV) solutions are employed, but in some cases the presence of
nitrate ion is undesirable. The nitrate ion may be removed by evaporating the
solid reagent with concentrated sulphuric acid, or alternatively a solution of the
nitrate may be precipitated with aqueous ammonia and the resulting cerium(IV)
hydroxide filtered off and dissolved in sulphuric acid.

Internal indicators suitable for use with cerium(IV) sulphate solutions include
N-phenylanthranilic acid, ferroin, and 5,6-dimethylferroin.

X, 108. PREPARATION OF O.IN CERIUM(IV) SULPHATE. Method
A. Dissolve about 28 g A.R. ammonium cerium(IV) nitrate (equivalent
= 548.23) in 100 cm^ water in a 600-cm^ beaker, add dilute ammonia solution
slowly and with stirring until a slight excess is present (about 60 cm^ ca. 2.51V-
ammonia solution are required). Filter the precipitated cerium(IV) hydroxide
with suction through a 7-cm sintered glass funnel, and wash with five 50-cm^
portions of water to remove ammonium nitrate; leave the precipitate ‘on the
water pump’ for about 30 minutes to remove as much water as possible. Transfer
the precipitate back to the original beaker as far as possible and remove the
residual hydroxide on the sintered glass filter by washing with four 50-cm*
portions of 2M-sulphuric acid previously warmed to about 60 °C. Add the
washings to the precipitate in the beaker, and warm until the precipitate dissolves
completely. Allow to cool, transfer the solution to a 500-cm^ graduated flask, and
make up to the mark with distilled water. The resulting solution of cerium(IV)
sulphate is about O.IJV, and requires standardisation before use.

Method B. Evaporate 55.0 g of A.R. ammonium cerium(IV) nitrate almost
to dryness with excess (48 cm^) of concentrated sulphuric acid in a Pyrex
evaporating-dish. Dissolve the resulting cerium(IV) sulphate in M-sulphuric acid
(28 cm^ concentrated sulphuric acid to 500 cm^ water), transfer to a 1-dm^
graduated flask, add M-sulphuric acid until near the graduation mark, and make
up to the mark with distilled water. Shake well.

Method C. The molecular weight and also the equivalent of cerium(IV)
sulphate Ce(S04)2 and ammonium cerium(IV) sulphate (NH4)4[Ce(S04)4],2H20
are 333.25 and 632.56 respectively.

Weigh out 35-36 g of pure cerium(IV) sulphate into a 600-cm^ beaker, add 56
cm of l:l-sulphuric acid and stir, with frequent additions of water and gentle
warming, until the salt is dissolved. Transfer to a 1-dm^ glass-stoppered graduated
flask and, when cold, dilute to the mark with distilled water. Shake well.

Alternatively, weigh out 64-66 g of ammonium cerium(IV) sulphate into a
solution prepared by adding 28 cm^ of concentrated sulphuric acid to 500 cm^ of
water: stir the mixture until the solid has dissolved. Transfer to a 1-dm^
graduated flask, and make up to the mark with distilled water.

Method D. Place about 21 g of cerium(IV) hydroxide in a 1500-cm^ beaker,
and add, with stirring, 100 cm^ of concentrated sulphuric acid. Continue the

365

\\ 109 QUANTITATIVnjNORGANIC ANALYSIS

stirring and introduce 300 cm^ of dislilicd water slowly and cautiously. Allow to
stand overnight, and if any residue remains, filter the solution into a 1-dra’
graduated flask and dilute to the mari;.

X, 109. STANDARDISATION OF CERIUM(IV) SULPHATE SOLU-
TIONS. Dheusskm. Tlie most trustworthy method for standardising
ceriumflV) sulphate solutions is with pure ar.scnicflD) oxide. Tiie reaction
between ccriumflV) sulphate solution and ar.scnicfll!) oxide is very slow at the
ordinary temperature; it is necessary to add a trace of osmium tciroxidc as
catalyst. The ur.senicllll) oxide is dissolved in sodium hydroxide solution, the
solution acidified witii dilute .sulphuric acid, and after adding 2 dropsofan'o.smic
acid' solution prepared by dissolving 0.! g osmium tetroxide in 40 cm^ O.LV-
sulpliuric acid, and the indicator H- 2 drops ferroin or 0,5 cm^ N-
phenylanthranilicacid), it is titrated with the ccriumflV) sulphate solution to the
first sharp colour change: orange-red to very p.'dc blue or yellowish-green to
purple respectively.

4 ll3As03-‘-H,0 - aeV' -f HjAsO^-i 2H*

Standardisation may also be carried out using pure iron, and also with A.R.
sodium oxalate; in this last ease, an indirect procedure must be used as the redox
indicators tire themselves oxidised at the elevated temperatures which are
ncccs,sary. 71ic procedure, therefore, is to add an excess orihcccriitni(IV)so!ution,
and then, after cooling, the e.xcess is determined by b;ick titmtion with an ironfll)
solution. It is possible to carry out a direct titration of the sorlium oxalate if a
potentiometric proa-durc is used (Chapter XIV).

Procedure A. Standardisation with arscnicflH) oxide. Weigh out
accurately about 0.2 g of A.R. arseniq'III) oxide, previously dried at 105-1 10 X
for 1-2 hours, and transfer toadCRI-cmMicaker or to a 350-cm^ conical flask. Add
20 cm-' of about 2A/-sodium hydroxide solution, and warm the mixture gently
until the arsciiiolll) oxide has compkiely dissolved. Coo! to room temperature,
and add 100 cm' water, followed by 25 cm-' 2.5.^f-,stI!ph^ric acid. Then add 3
drops O.OlA.f-osmium tetroxide solution (0.25 g osmium tetroxide di.ssolved in
100 cm' 0.05AI-sulphtiric acid) and 0,5 cm' lY-phcnylanthranilic acid indicator
(or 1 -2drajis of ferroin). Titrate svith the 0.1;V-cerium(l V) .sulphate solution until
the first sharp colour change occurs (see Disch.s.m'ou above). Repeat with tsvo other
samplc.s of approximately equal weight of arscnicdll) oxide.

Procedvre B. Standardisation with pure iron. Weigh out accurately 0.15-
0.20 g A.R, iron wire, and then proceed exactly os described in Section X. 92,
Prni'cdwc C. I iirate the resulting solution with the ccriumflA') sulphate solution,
using any of the indicators referred to in the Diseussinn.

Procedure C. Sland.nrdisation with sodium oxalate. Prepare an
approximately 0. 1 N solution of ammonium ironlD) sulphate in dilute sulphuric
acid and titrate svith the cerium(lV) sulphate .solution using ferroin indicator.

Weigh out accurately about 0,2 g A.R, sodium oxalate into a 250-cm' conical
flask and add 25 -30 cm' Af-sulphuric acid. Heat the solution to about 60 ‘'C and
then add about 30 cm' of the ccrium(! V) solution to be standardised dropssisc.
adding the solution as rapidly as possible consistent with drop formation. Reheat
the solution to (\0 C, and then add a further 10 cm' of the ccriumflV) solution
Allow to stand for three minutes, then cool and back-tilratc the excess ccriumflV)
with tile ironfll) solution using ferroin as indicator.

366

TITRIMETRIC ANALYSIS X, 110/111

Practically, all the determinations described under potassium permanganate
and potassium dichromate may be carried out with cerium(IV) sulphate. Use is
made of the various indicators already detailed and also, in some cases where
great accuracy is not required, of the pale yellow colour produced by the
cerium(IV) sulphate itself. Only a few determinations will therefore be considered
in some detail.

X, 110. DETERMINATION OF COPPER. Discussion. Divalent copper is
quantitatively reduced in 2/V-hydrochloric acid solution by means of the silver
reductor (Section X, 145) to the copper(I) state. The solution of the copp,er(I) salt
is collected in a solution of ammonium iron(III) sulphate, and the_ Fe^ ^ ion
formed is titrated with standard cerium(IV) sulphate solution using ferroin or N-
phenylanthranilic acid as indicator.

Comparatively large amounts of nitric acid, and also zinc, cadmium, bismuth,
tin, and arsenate have no effect upon the determination; the method may
therefore be applied to determine copper in brass.

Procedure (copper in crystallised copper sulphate). Weigh out accurately
about 3.1 g A.R. copper sulphate crystals, dissolve in water, and make up to 250
cm^ in a graduated flask. Shake well. Pipette 50 cm^ of this solution into a small
beaker, add an equal volume of ca. 4M-hydrochloric acid. Pass this solution
through a silver reductor at the rate of 25 cm^ per minute, and collect the filtrate
in a 350-cra^ conical flask charged with 20 cm^ 0.5M-iron(III) alum solution
(prepared by dissolving the appropriate quantity of A.R. iron(III) alum in 0.5M-
sulphuric acid). Wash the reductor column with six 25-cm^ portions of 2M-
hydrochloric acid. Add 1 drop of ferroin indicator or 0.5 cm^ JV-
phenylanthranilic acid, and titrate with O.lN-cerium(IV) sulphate solution. The
end-point is sharp, and the colour imparted by the Cu^"'" ions does not interfere
with the detection of the equivalence point.

Procedure (copper in copper(I) chloride). Prepare an ammonium iron(III)
sulphate solution by dissolving 10.0 g of the A.R. salt in about 80 cm^ of 6JV-
sulphuric acid and dilute to 100 cm^ with acid of the same strength. Weigh out
accurately about 0.3 g of the sample of copper(I) chloride into a dry 250-cm^
conical flask and add 25.0 cm^ of the iron(III) solution. Swirl the contents of the
flask until the copper(I) chloride dissolves, add a drop of two of ferroin indicator,
and titrate with standard 0. liV-cerium(IV) sulphate.

Repeat the titration with 25.0 cm^ of the iron solution, omitting the addition of
the copper(I) chloride. The difference in the two titrations gives the volume of
O.lN-cerium(IV) sulphate which has reacted with the known weight of copper(I)
chloride.

X,lll. DETERMINATION OF MOLYBDATE. Discussion. Molybdates
[Mo(VI)] are quantitatively reduced in 2M-hydrochloric acid solution at 60-80
°C by the silver reductor to quinquevalent molybdenum [Mo(V)]. The reduced
molybdenum solution is sufficiently stable over short periods of time in air to be
titrated with standard cerium(IV) sulphate solution using ferroin or N-
phenylanthranilic acid as indicator. Nitric acid must be completely absent; the
presence of a little phosphoric acid during the reduction of the molybdenum(VI)
IS not harmful and, indeed, appears to increase the rapidity of the subsequent
oxidation with cerium(IV) sulphate. Elements such as iron, copper, and

367

X. 112/113 quantitative INORGANIC ANALYSIS

vanadium interfere; nitrate interferes, since its reduction is catalysed by the
presence of molybdates.

procedure. Weigh out accurately about 2.5 g A,R. ammonium molybdate
(NHj)f,Mo-024,4H,0. dissolve in water and make up to 250 cm^ in a graduated
flask. Pipette 50 cm*’ of this solution into a small beaker, add an equal volume of
4A/-hydrochloric acid, then .3 cm*' of 85 per cent phosphoric acid, and heat the
solution to 60 SO ' C. Pour hot 2AMiydrochloric acid through a silver rcductor,
and then pass the molybdate solution through the hot rcductor at the rate of
about 10 cm*' per minute. Collect the reduced solution in a 4CK)-cm ’ beaker or
35D-cm-' conical flask, and wash the rcductor with six 25-cm-' ponions of 2M-
hydrochloric acid : the lirst two washings should he made with the hot acid (rate:
10 cm^ per minute) and the hast four wa.shings willi the cold acid (rate; 20-25 cm^
per minute). Cool the .solution, add one drop of ferroin or 0.5 cm^ N~
phenylanthranilic acid, and titrate witit .standard O.I A'*cerium(lV) sulphate. The
precipitate of cerium(IV) phosphate, which is initially formed, dissolves on
shaking, Add the last 0.5 cm-' of the reagent dropwisc and with vigorous stirring
or shaking.

X, 112. ni-TERMINATION OF TFI.LURin-. A measured e.xce.ss of
standard O.lA’-ceriumflV) sulphate is atided to the iellurium{lV) solution (200
cm*') containing 10 cm ’ of concentrated hydrochloric acid and about 0.05 g of
chromium(IIl) sulphate as catalyst.* The solution is boiled for 10 minutes, then
cooled and back- titrated with standard O.I iV-ammoniutn iron(ll) sulphate, using
iV-phcnylanthranilic acid or ferroin as indicator. The tellurium is oxidised from
the tetra- to the hexavalcnl stage. Selenium doe.s not interfere.

X,11.3. DtriTCRMlNATION OF CFRlUMdU). Method A. TTic

ccriunidlDsali in the form of sulphate in 100 cm' of 1 ;4-sulphuricacid is treated
with 2 g of ammonium .sulphate. 1 g of A.R. sodium bismuthate is added, and the
solution heated to boiling. The mixture is cooled somewhat, 50 cm*’ of 2 perami
sulphuric acid added, filtered through a Gooch, porcelain, or sintered glass
liltcring crucible, and the crucible washed with 10(>-150 cm’ of 2 per cent
sulpluiric acid.

2Cc’ * + DiOj - -f- 61 r 2Cc* ' f Ri’ ’ -f 3H -O

Excess of0,025A'-ammonium iron(Il) sulphate is added (as shown by the chance
from yellow to colourless, and the con.scqucnt complete reduction of the
ccriuni(lV) to the cerium(in) salt), and the excess of iron(ll) salt titrated with
0. 1 A'-potassium permanganate to the first appearance of a pink colour.

Method B. 100-300 cm’ of the solution, containing 0.1 -0.3 g of cerium and
2.5-1. S cm’ of concentrated .sulphuric acid, arc treated with 1-1.5 g of A.R.
ammonium persulphate and 10 drops of 0.1 Af-si!vcr nitrate solution (catalyst),
and then boiled for 10 minutes. The .solution is cooled to room temperature, and
is ready for titration. Two procedures may he used.

Ccnimi(IVj sulplialc alone Ones noi o\idiw selciiiic or icllurite. but il oxidises chromium to the
Ilex.*! valent slate and this, in Iiirn. oxidises Iclluriletbut not selenite) to ibcbcxav.iIcnlcondition:any
chromiurni VI) at the end is rcdiiccrl bv the iron(II) siilpbaic. lltc ccritimflV) ion acts a.s a polentul
mcdiaior (compare Section X, 3 . 1 . A]

368

TITRIMETRIC ANALYSIS X, 114

(i) Add 10-20 cm^ of 10 per cent potassium iodide solution, and titrate the

liberated iodine with O.IN- or O.oiSiV^-sodium thiosulphate. To avoid the
oxidation of the hydriodic acid by the air, the titration should be performed in an
inert atmosphere (N2 or GO2). , .

(ii) Titrate the solution with O.IN- or 0.025N-ammonium iron(II) sulphate,
using N-phenylanthranilic acid or ferroin as indicator. . .

X, 114. DETERMINATION OF NITRITES. Discussion. Satisfactory
results are obtained by adding the nitrite solution to excess of standard O.IN-
cerium(IV) sulphate, and determining the excess of cerium(IV) sulphate with a
standard iron(II) solution (compare Section X, 97).

2Ce''+ + NO2 " + H2O = 2Ce3 + -H NO3 - + 2H+

For practice, determine the percentage of NOj in potassium nitrite, pr the
purity of sodium nitrite, preferably of A.R. quality.

Procedure. Weigh out accurately about 1.5 g of sodium nitrite and dissolve
it in 500 cm^ of boiled-out water in a graduated flask. Shake thoroughly. Place 50
cm^ of standard 0.1N-cermm(IV) sulphate in a conical flask, and add 10 cm^ of
2M-sulphuric acid. Transfer 25 cm^ of the nitrite solution to this flask by means of
a pipette, and keep the tip of the pipette below the surface of the liquid during the
addition. Allow to stand for 5 minutes, and titrate the excess of cerium(IV)
sulphate with standard O.lN-ammonium iron(II) sulphate, using ferroin or N-
phenylanthranilic acid as indicator. Repeat the titration with two further
portions of the nitrite solution. Standardise the iron solution by titrating 25 cm^
of it with the cerium(IV) solution in the presence of dilute sulphuric acid.

Determine the volume of the standard cerium(IV) sulphate solution which has
reacted with the nitrite solution, and therefrom calculate the purity of the sodium
nitrite employed.

Note. Cerium(IV) sulphate may also be used for the following analyses.

1. Hydrogen peroxide. The diluted solution, which may contain nitric,
sulphuric, or hydrochloric acid in any concentration between 0.5 and 3N, is
titrated directly with standard cerium(IV) sulphate solution, using ferroin or N-
phenylanthranilic acid as indicator. The reaction is :

2Ce‘^ + + H2O2 = 2Ce3 + -b O2 -f 2H +

2. Persulphate (peroxydisulphate). Persulphate cannot be deterrhined
directly by reduction with iron(II) because the reaction is too slow:

S208^“+2Fe2+ =28042" -b2Fe3 +

An excess of a standard solution of iron(II) must therefore be added and the
excess back-titrated with standard cerium(IV) sulphate solution. Erratic results
are obtained, depending upon the exact experimental conditions, because of
induced reactions leading to oxidation by air of iron(II) ion or to decomposition
of the persulphate; these induced reactions are inhibited by bromide ion in
concentrations not exceeding IM and, under these conditions, the determination
may be carried out in the presence of organic matter.

To 25.0 cm^ of 0.01-0.01 5M-persulphate solution in a 150-cm2 conical flask,
add 7 cm^ of 5M-sodium bromide solution and 2 cm^ of 3M-sulphuric acid.
Stopper the flask. Swirl the contents, then add excess of 0.05N-ammonium
iron(II) sulphate (15.0 cm^), and allow to stand for 20 minutes. Add 1 cm^ of

369

X. 115 QUANTITATIVn INORGANIC ANALYSIS

0.001 Af-fcrroin indicator, and titrate the excess of Fc^" ion with 0,02/Y.
ceriuni(IVj sulpliatc in 0.5A/-sulphiiric acid to the first colour change from
orange to vcliovv.

3. Urnniiim. Uranium, as uranyl sulphate in solutions 4M in hydrochloric
acid, is reduced quantitatively to tctravalcnt uranium on pa.s',sagc through a silver
redu’etor at 60-90 ' C: the uraniumll V) can he titrated with standard cerium(lV)
sulphate solution.

Dissolve the uranium .salt, containing 0,1-0,3 g of uranium, in 50 cm’ of AM-
hydrochloric acid and heat to 60-90 ' C. Pre-treat a silver rcducior (Section X,
145) with hot 4 A/-hydrochioric acid and pass the uranium! VI) solution throughit
at a rate of 20 cm’ per minute. Wash with hot 4Af-liydrochlor!C acid. Coo! the
reduced .solution, acid 3 cm' of 85 per txmt phosphoric acid and one drop of
ferroin indicator. Titrate with standard O.UV-cerium(lV) sulphate to the
disappearance of the pink colour. .A little silver chloride m.ay precipitate, but this
docs not alTcct the analysis. Run a blank determination and subtract the value
found from the litre found in the uranium titration.

Calculate the percentage of uranium in the sample.

U-*" ■f2Ce'’‘ 4-211,0 =• UOw' -!-2CV* 4..4|r

4. Iroa Tlie determination of iron (c.g., in an iron ore) c,an he carried out
by following the procedure given in Section 109, Procedure B,

5. 0.valate,s. Oxalates can lx* determined by means of the indirect method
described in Section X. 109, Procedure C.

6. Hcxacyanofcrratedl) can he determined by titration in M- 11 ,S 04 using
.V-phenylanthranilic acid.

Oxidation and Reduction pr()ccs.sc.s involving iodine
lodonietric titrations

X, 115. GFNERAL DLSCUSSION. The direct iodomctric lilr,ation method
(sometimes termed iodinwtry) refers to titrations with a standard solution of
iodine. The indirect iodomctric (itr.ation method (sometimes termed iodometry)
deals wilii the titration u/' iodine liberated in chemical reaction.s. The normal
reduction potential of the reversible system;

I, (soiid)-t-2ceit2I '

is 0.5345 volt. The above equation refers to a s.ituratcd aqueous solution in the
presence of solid iodine; this half-cell reaction will occur, for example, towards
the end of n titration of iodide with an oxidising agent such as potassium
permanganate, when the iodide ion concentration becomes relatively low. Near
the beginning, or in most iodomctric titrations, when an excess of iodide ion is
present, the tri-iodide ion is formed

ij(aq.)-fr eii.r

since iodine is readily soluble in a solution of iodide. The half-cell reaction is
belter written:

! j ■ + 2e 31 "

and the standard reduction potential is 0.5355 volt. Iodine or the tri-iodide ion
is tliercfore a much weaker oxidising agent than potassium permanganate,
potassium dichromate, and cenum(fV) sulphate.

370

TITRIMETRIC ANALYSIS X, 115

In most direct titrations with iodine (iodimetry) a solution of iodine in
potassium iodide is employed, and the reactive species is therefore the tri-iodide
ion 13 “. Strictly speaking, all equations involving reactions of iodine should be
written with I 3 “ rather than with Ij, e.g.,

I3- + 2 S 2 O 32 - = 3I:-|-S4062- ■ ,

is more accurate than . ,

.13+28303"- =21-4- 8406=*-

For the sake of simplicity, however, the equations' in this book will usually be
written in terms of molecular iodine rather than the tri-iodide ion.

Strong reducing agents (substances with a much lower reduction potential),
such as tin(II) chloride, sulphurous acid, hydrogen sulphide, and sodium
thiosulphate, react completely and rapidly with iodine even in acid solution. With
somewhat weaker reducing agents, e.g., trivalent arsenic, or trivalent antimony,
complete reaction occurs only when the solution is kept neutral or very faintly
acid; under these conditions the reduction potential of the reducing agent is a
minimum, or its reducing power is a maximum.

If a strong oxidising agent is treated in neutral or (more usually) acid solution
with a large excess of iodide ion, the latter reacts as a reducing agent and the
oxidant will be quantitatively reduced. In such cases, an equivalent amount of
iodine is liberated, and is then titrated with a standard solution of a reducing
agent, which is usually sodium thiosulphate.

The normal reduction potential of the iodine-iodide system is independent of
the pH of the solution so long as the latter is less than about 8 ; at higher values
iodine reacts with hydroxide ions to form iodide and the extremely unstable
hypoiodite, the latter being transformed rapidly into iodate and iodide by self-,
oxidation and reduction :

I 2 + 2 OH- =r +IO-+H 3 O

310- =2I-+I03-

The reduction potentials of certain substances increase considerably, with
increasing hydrogen-ion concentration of the solution. This is the case with
systems containing permanganate, dichromate, arsenate, antimonate, bromate;
etc., i.e., with anions which contain oxygen and therefore require hydrogen for
complete reduction. Many weak oxidising anions are completely reduced by
iodide ions if their reduction potentials are raised considerably by the presence in
solution of a large amount of acid.

By suitable control of the pH of the solution, it is sometimes possible to titrate
the reduced form of a substance with iodine, and the oxidised form, after the
addition of iodide, with sodium thiosulphate. Thus with the arsenite-arsenate
system;

H 3 ASO 3 + 12 + H 3 O ^H3As04 + 2H+ + 21-

the reaction is completely reversible. At pH values between 4 and 9, arsenite can
be titrated with iodine solution. In strongly acid solutions, however,' arsenate is
reduced to arsenite and iodine is liberated. Upon titration' ’with sodium
thiosulphate solution, the iodine is removed and the reaction proceeds from right

371

X, 116 QUANTITATIVE INORGANIC ANALYSIS

Two important sources of error in titrations involvins iodine arc: (a) loss of
iodine owing to its appreciable volatility: and (h) acid solutions of iodide are
oxidised by oxygen from Ibe air:

4r+Oj + 4H" ==21j+2HjO

In the presence of excess of iodide, the volatility is decreased markedly through
the formation of the tri-iodide ion; at room temperature the loss of iodine by
volatilisation from a solution containing at least 4 per cent of potassium iodide is
negligible provided the titration is not prolonged unduly. Titrations should be
performed in cold solutions in conical flasks and not in open beakers. Ifa solution
is to stand it should be kept in a plass-stopjvercd vessel. The atmospheric
oxidation ofioditle is negligible in neutral solution in the absence ofcalalysts. but
the rate of oxidation increases rapidly witli decreasing pH. llic reaction is
catalysed by certain metal ions of variable valency (particularly copper), by nitrite
ion, and also by strong light. For this reason titrations should not lx; performed in
direct sunlight, and solutions containing iodide should lx: stored in amber glass
bottles. Furthermore, the air oxidation of iodide ion may be induced by the
reaction between iodide and the oxidising agenu especially when the main
reaction is slow. Solutions conttiining an excess of iodide and acid must therefore
not be allowed to stand longer than nccc.ss3ry before titration of the iodine. If
prolonged standing is ncccss;iry (as in the titration of vanadate or Fc^ * ions) the
solution should be free from air before the addition of iodide and the air di-splaccd
from the titration vessel by c.irhon dioxide (c.g,. by adding small portions (().2-0.5
g) of pure sodium hydrogcncarbonate to the acid solution, or a iillle Dry lee);
potassium iodide is then introduced and the glass stopper replaced immediately.

It seems appropriate to refer at this point to the uses of a standard solution
containing potassium iodide and potassium iodntc. Tltis solution is quite stablcand
yields iodine when treated with acid:

-fbir =31:-(-3H:0

The standard soUitioti is prepared b\ dissolving a weighed amount of pure
potassium iodatc in a solution containing a slight excess of pure potassium
iodide, and diluting to a definite volume. This solution has two important uses.
The first is as a source of a known quantity of iodine in titrations (compare
Section X, 1 ISA); it must be added to a solution containing strong acid: it cannot
be employed in a medium which is neutral or possc.sscs a low acidity.

The second use is in the determination of the add content of solutions
iodometrically or in the .standardi.sathei of .solutions of strong acids. It is evident
from the above equation that the amount of iodine liberated is equivalent to the
acid content of the solution. Thus if, say, 25 cm-* of an approximately O.LV
solution of a strong acid is treated with a slight c.xct‘ss of pola,ssium iodatc (say, 30
cm-’ of O.liV-potassium iodatc solution. Section X. 132) and a slight excess of
potassium iodide solution (say, 1 0 of a 10 per cent solution), and the liberated
iodine titrated with standard 0.1/V-.SQciium thiosulphate with the aid of starch as
an indicator, the normality of the acid may be readily evaluated.

X, 116. DCTECTfON OF THE END-POINT. A solution of iodine in
aqueous iodide has an intense yellow to brown colour. One drop of O.UV-iodinc
solution imparts a perceptible pule yellow colour to 100 enr’ of water, so that in
otherwise colourless solutions iodine can serve as its own indicator. The test is

372

TITRIMETRIC ANALYSIS X, 116

made much more sensitive by the use of a solution of starch as indicator; Starch
reacts with iod ine in the presence of iodid e to fnrm-an4ntensely blue-coloured
complex, which is visible at very low^coMentratiom of iodine. The sensitivity of
the colour rSctibn^is such thaF'’a blue colour is visible when the iodine
concentration is 2xlO“^M and the iodide concentration is greater than 4
X at 20 °C. The colour- sensitivity decreases with increasing temperature

of the solution; thus at 50 °C it is about ten times less sensitive than at 25 °C. The
sensitivity decreases upon the addition of solvents, such as ethanol: no colour is
obtained in solutions containing 50 per cent ethanol or more. It cannot be used in
a strongly acid medium because hydrolysis of the starch occurs.

Starches can be separated into two major components, amylose and
amylopectin, which exist in different proportions in various plants. Amylose,
which is a straight-chain compound and is abundant in potato starch, gives a blue
colour with iodine and the chain assumes a spiral form. Amylopectin, which has a
hranched-chain structure, forms a red-purple product, probably by adsorption.

The great merit of starch is that it is inexpensive. It possesses the following
disadvantages: (i) insolubility in cold water; (ii) instability of suspensions in
water; (iii) it gives a water-insoluble complex with iodine, the formation of which
precludes the addition of the indicator early in the titration (for this reason, in
titrations of iodine, the starch solution should not be added until just prior to the
end-point when the colour begins to fade); and (iv) there is sometimes a ‘drift’ end-
point, which is marked when the solutions are dilute.

Most of the shortcomings of starch as an indicator are absent in sodium starch
glycoUate. This is a white, non-hygroscopic powder, readily soluble in hot water
to give a faintly opalescent solution, which is stable for many months; it does not
form a water-insoluble complex with iodine, and hence the indicator may be
added at any stage of the reaction. With excess of iodine (e.g., at the beginning of a
titration with sodium thiosulphate) the colour of the solution containing 1 cm^ of
the indicator (0.1 per cent aqueous solution) is green; as the iodine concentration
diminishes the colour changes to blue, which becomes intense just before the end-
point is reached. The end-point is very sharp and reproducible and there is no
‘drift’ in dilute solution.

Carbon tetrachloride has been used in certain reactions instead of starch

solution. One dm^ of water at 25 °C will dissolve 0.335 g of iodine, but the same
volume of carbon tetrachloride will dissolve about 28.5 g. Iodine is therefore
about eighty-five times as soluble in carbon tetrachloride as it is in water, and the

carbon tetrachloride solution is highly coloured. When a little carbon
tetrachloride is added to an aqueous solution containing iodine and the solution
well shaken, the great part of the iodine will dissolve in the carbon tetrachloride;
the latter will fall to the bottom since it is immiscible with water, and the colour of
the organic layer will be much deeper than that of the original aqueous solution.
The reddish-violet colour of iodine in carbon tetrachloride is visible in very low
concentrations of iodine ; thus on shaking 10 cm^ of carbon tetrachloride with 50
cm of 2 X 10“^ N-iodine, a distinct violet coloration is produced in the organic
layer. This enables many iodometric determinations to be carried out with

comparative ease. The titrations are performed in 250-cm^ glass-stoppered
bottles or flasks with accurately ground stoppers. After adding the excess of
potassium iodide solution and 5-10 cm^ of carbon tetrachloride to the reaction
mixture, the titration with sodium thiosulphate is commenced. At first the
presence of iodine in the aqueous solution will be apparent, and gentle rotation of

373

X, 117 OUANTITATIVK INORGANIC ANALYSIS

the liquid causes sufficient mixinp. Towards the end of the titration the bottle or
flask is stoppered and shaken after each addition of sodium thiosulphate
solution: the end-point is reached when the carbon tetrachloride just becomes
colourless. Equally satisfactory results can l>c obtained with chloroform.

Preparation and use of starch solution. Make a paste of 1.0 g of soluble
starch with a little water, and pour the paste, with constant stirring, into 100 cm^
of boiling water, and boil for 1 minute. Allow the solution to cool and add 2-3 gof
potassium iodide. Keep the .solution in a stoppered bottle.

Only freshly prepared starch solution .should be used. T wo cm^ of a 1 per cent
solution per 100 cm ' of the solution to be titrated is a satisfactory amount; the
same volume of starch solution should always lx: added in a titration. In the
titration of iodine, starch must not be added until just before the end-point is
reached. Apart from the fact that the fading of the iodine colour is a good
indication of the approach of the end-point, if the starch solution is added when
the iodine concentration is high, some iodine may remain adsorbed even at the
end-point. The indicator blank is negligibly small in iodimeiric and iodomctric
titrations ofO.liV-solutions; with more dilute solutions.it must be determined in
a liquid having the same composition as the solution titrated has at the end-point.

A solid solution of starch in urea may also be employed. Reflux 1 g soluble
starch and 19 g urea with xylene. .At the boiling point of the organic .solvent the
urea mclt-s with little decomposition, and the .starch dissolves in the molten urea,
Allow to cool, then rc.movc the solid mass and powder it: store the product in a
stoppered bottle. A few milligrams of this .solid added to an aqueou-s .solution
containing iodine then bchases like the usual starch indituitor.

Preparation and use of sodium starch Rlycollate indicator. Sodium starch
glycollate, prepared as described belosv. dissolves slowly in cold but rapidly in hot
svatcr. It is b.M dissolved by mixing, sas, 5.0 gof the finds posvdered solid with 1-
2 cm' ethanol, adding 100 end cold water, and boiling for a few minutes with
vigorous stirring; a faintly opalescent solution results. Tlus 5 per amt slock
solution is diluted to I per cent strength as required. The most convenient
concentration for use as an indicator is 0.1 mg/cm" ■*, i,c.. 1 cni^ of the 1 percent
aqueous solution is added to 100 cm ' of the solution being titrated,

X, 117. PREPARATION OF 0.1 A’-SOI)l 1 IM 1 HlOSULPIl.ATf; Discus-
Sinn. Sodium thiosulphate Na.S.Q, .511,0 is readily obtainable in a state of
high purity, but there is always some uncertainty as to the exact water content
because of the cfilorcscent nature of the s.dt and for other reasons. Tltc substance
is therefore unsuitable as a primary -tandard. It is a reducing agent by virtue of
the half-cell reaction:

2.S,0/-ae S,0,= - -f2r

the equivalent of sodium thiosulphate pentahydrate is the tnolc. or 24S.1S. An
approximately 0.1 A' solution is picparcd by dissolving about 25 g A.R.
crystallised sodium thiosulphate in 1 litre of water in a graduated flask. The
solution is standardised by any of the nictliod.s described below.

Before dealing with these, it is nece.ssary to refer briefly to the stability of
thiosulphate solutions. .Solutions prepared with conductivity (equilibrium) water
arc perfectly stable. However, ordinary distilled water usually contains an excess
of carbon dioxide; this may c;uise a slow decomposition to take place with the
formation of sulphur;

37-1

TITRIMETRIC ANALYSIS X, 118

S2032“+H+=HS03~+S

Moreover, decomposition may also be caused by bacterial action (e.g.,
thiobacillus thioparus), particularly if the solution has been standing for some
time. For these reasons, the following recommendations are made ;

1. Prepare the solution with recently boiled distilled water.

2. Add 3 drops of chloroform or 10 mg of mercury{II) iodide per litre; these
compounds improve the keeping qualities of the solution.

(Bacterial activity is least when the pH lies between 9 and 10. The addition of a
small amount, 0.1 g per litre, of sodium carbonate is advantageous to ensure the
correct pH. In general, alkali hydroxides, sodium carbonate (>0.1 g/1), and
sodium tetraborate should not be added, since they tend to accelerate the
decomposition;

8303''“ +2O2 + H3O ^23042- +2H+)

3. Avoid exposure to light, as this tends to hasten the decomposition.

The standardisation of thiosulphate solutions may be effected with potassium
iodate, potassium dichromate, copper and iodine as primary standards, or with
potassium permanganate or cerium(IV) sulphate as secondary standards. Owing
to the volatility of iodine and the difficulty of preparation of perfectly pure iodine,
this method is not a suitable one for beginners. If, however, a standard solution of
iodine (see Sections X, 119, 120) is available, this may be used for the
standardisation of thiosulphate solutions.

Procedure. Weigh out 25 g A.R. sodium thiosulphate crystals, Na2S203,
5H2O, dissolve in boiled-out distilled water, and make up to 1 litre in a graduated
flask with boiled-out water. If the solution is to be kept for more than a few days,
add 0.1 g sodium carbonate or 3 drops of chloroform.

X, 118. STANDARDISATION OF SODIUM THIOSULPHATE SOLU-
TIONS. A. With potassium iodate. A.R. potassium iodate has a purity of at
least 99.9 per cent: it can be dried at 120 °C. This reacts with potassium iodide in
acid solution to liberate iodine:

I03-+5I--f6H+ =3l2-f-3H20

Its equivalent as an oxidising agent is 5 mole or 214.00/6; a O.IJV solution
therefore contains 3.567 g of potassium iodate per dm^.

Weigh out accurately 0. 14-0. 1 5 g of pure dry potassium iodate, dissolve it in 25
cm^ of cold, boiled-out distilled water, add 2 g of iodate-free potassimn iodide*
and 5 cm^ of M-sulphuric acid (1). Titrate the liberated iodine with the
thiosulphate solution with constant shaking. When the colour of the liquid has
become a pale yellow, dilute to ca. 200 cm^ with distilled water, add 2 cm^ of
starch solution, and continue the titration until the colour changes from blue to
colourless. Repeat with two other similar portions of potassium iodate.

Note. 1. Potassium iodate has a small equivalent (35.67) so that the error
in weighing 0.14-0.15 g may be appreciable. In this case it is better to weigh out
accurately 3.567 g of the A.R. salt (if a slightly different weight is used, the exact
normality is calculated), dissolve it in water, and make up to 1 dm^ in a graduated
flask. Twenty-five cm^ of this solution are treated with excess of pure potassium

The absence of iodate is indicated by adding dilute sulphuric acid when no immediate yellow
coloration should be obtained. If starch is added, no immediate blue coloration should be produced.

375

X, 118 OUANTITATIVK INORGANIC ANALYSIS

iodide (1 g of the solid or 10 cin^ of 10 per cent solution), followed by 3 an^ of Af.
sulpliuric acid, and the liberated iodine is titrated as detailed above.

B. Witfi potassium dichromate. Potassium dichromate is reduced by an
acid solution of potassium iodide, and iodine is .set free:

Cr,0,^- -i-dr + 14H • - 2Cr’^ +3[,+7U,0

This reaction is subject toa number of errors:(I)thc hydriodic acid (from excess
of iodide and acid) is readily oxidised by air, especially in the presence of
chromiumdli) salts, and (21 it is not instantaneous. U is itccoidingly best to passa
current of carbon dio.vide through the reaction Dask before and during the
titration (a more convenient but less cDicieni method is to add .some solid sodium
hvdrogcncarbonatc to the acid solution, and to keep the flask covered as much
as possible), and to .allow 5 minutes for its completion.

Place 100 cm' of cold, recently boiled distillcii water in a 5f)0-cm' conical,
preferably glass-.stoppercd. flask, add 3 g of iodatc-free potassium iodide and 2 g
of pure sodium hydrogencarhonate. and shake until the salts dissolve. Add 6 cm'
of concentrated bydrocliloric acid slowly whilst gently rotating the flask in order
to mix the liquids; nm in 25.0 cm' of standard O.JA'-pot:iSsium dichromatc(l),
mi.x the solutions well, .and wash I he .sides of the flask with a little boiled-out water
from the wash bottle. Stopper the fl.ask (or estver it with a small watch glass), and
allow to stand in tlie dark for 5 minute^ in order to complete the reaction. Rinse
the stopper or watch glass, and dilute tlie .solution with .''00 cm' of cold, boiled-
out water. Titrate the liberated ioilinc with the sodiimt thinsulphntc .solution
contained in a inirctte, whilst constantly rotating the liquid so as to thoroughly
mix the solutions. When most of the iodine has reacted as indicated by the
solution acquiring a yellowish-green colour, add 2 cm' of starch solution and
rinse down the side.s of tlie flask ; the colour should change to blue. Continue the
addition of the tbiosulpliatc solution dropwisc. and swirling the liquid
constantly, until 1 drop cliange- the colour from grcenish-Wuc to light green. The
end-point is .sitarp. and is readily observed in a good light against a white
background. Carry out a blank determinatton, substituting distilled water for (he
potassium dichroniatc solution; if the potassuim iodide is iodatc-free, this should
be negligible.

Note. 1, If preferred, about 0,20 g of A.R. potassium dichromatc may be
accurately weighed out, dissohed in 50' cm' of cold, boiled-out water, and the
titration carried out as detailed tibove.

The following (iltcrnative procedure uiilisc.s :t trace of copper s'ulphatc as a
c.-italysf to increase the speed of the reaction; in consequence, a weaker acid
(acetic acid) may Ik employed and the c.stcnt of atmospheric oxidation of
hydriodic acid reduced, f’lace 25.0 cm' of (khV-potassium dichromatc in a 250-
cm' conical flask, add 5.0 cm' of glacial acetic acid. 5 cm' of 0.001 A /-copper
.sulphate, and wash the sides of the flask witii distilled w ater. Add 30 cm' of 10 per
cent potassium iodide solution, and titrate the iodine as libenitcd with the
approxim.afcly 0. LV-thiosuiphalc solution, introducinc a little starch indicator
towards the end. TItc titration may be completed in 3-4 minutes after the
addition of the potassium iodide sohition. Subtract 0.05 cm' to allow for the
iodine liberated by the copper sulphate catalyst,

A .st.nndardised solution of potassium permanganate may be used in place of the
potassium dichromatc .solution, adding 2 cm' of concentrated hydrochloric acid

376

TITRIMETRIC ANALYSIS X, 119

to each 25-cm^ portion of potassium permanganate solution; in this case the
alternative procedure of weighing out a portion of the salt cannot be used.

C. With a standard solution of iodine. If a standard solution of iodine is
available (see Section X, 119), this may be used to standardise the thiosulphate
solution. Measure a 25.0-cm^ portion of the standard iodine solution into a 250-
cm^ conical flask, add about 150 cm^ distilled water and titrate with the
thiosulphate solution, adding 2 cm^ of starch solution when the liquid is pale
yellow in colour.

When thiosulphate solution is added to a solution containing iodine the
overall reaction, which occurs rapidly and stoichiometrically under the usual
experimental conditions (pH < 5), is;

2S2O32--H2 = S 406 '=--(- 2 I-

or 28203^- -t- 13- = +31"

It has been shown that the colourless intermediate 8203!“ is formed by a rapid
reversible reaction :

82O3"- +12^8203!- +1-

The intermediate reacts with thiosulphate ion to provide the main course of the
overall reaction:

S2O3I- +82O3"- = +r

The intermediate also reacts with iodide ion :

2S2O3I--H- = 84O62-+I3-;

this explains the reappearance of iodine after the end-point in the titration of very
dilute iodine solutions by thiosulphate.

D. With cerium(IV) sulphate. This method for standardising sodium
thiosulphate solutions makes use of a secondary standard, but gives satisfactory
results provided the experimental conditions given below are rigidly adhered to;
this is due to the fact that cerium(IV) sulphate solution contains free acid, which
may otherwise lead to appreciable errors.

For O.lA-cerium(IV) sulphate (8ections X, 108-109), use 25.0 cm^ of the ca.
O.liV-sodium thiosulphate solution, 0.3-0.4 g of pure potassium iodide, 2 cm^ of
0.2 per cent starch solution, dilute to 250 cm^, and titrate with the cerium(IV)
sulphate solution to the starch-iodine end-point, i.e., to a first permanent blue
colour.

The reaction is:

2Ce'^-*--f2I- = 2Ce3+-hl2

X, 119. PREPARATION OF O.IM-IODINE SOLUTION. Discussion.
0.335 gram of iodine dissolves in 1 dm^ of water at 25 "C. In addition to this small
solubility, aqueous solutions of iodine have an appreciable vapour pressure of
iodine, and therefore decrease slightly" in concentration on account of
volatilisation when handled. Both difficulties are overcome by dissolving the
iodine in an aqueous solution of potassium iodide. Iodine dissolves readily in
aqueous potassium iodide, the more concentrated the solution, the greater is the
solubility of the iodine. The increased solubility is due to the formation of a tri-
iodide ion:

I2 + I--I 3 -

377

X, 120 QUANTITATIVE INORGANIC ANALYSIS

The resulting holuiion has a much lower vapour pressure than a solution of iodine
in pure water, consequently the loss by volatilisation is considerably diminished.
Nevertheless, the vapour pressure is still appreciable so that prccciutiom should
always be taken (o keep vessels c<}nfnttjii!.!; iodine closed except during the actual
tiinttums. When an iodide solution of iodine is titrated with a rcductant, the free
iodine reacts with the reducing agent, this displaces the equilibrium to the left,
and eventually all the tri-iodide is decomposed ; the solution therefore behaves as
though it were a solution of free iodine.

For the preparation of standard iodine .solutions, A.R. or rcsublimed iodine
and iodatc-free {c.p.. A.R.) potassium iodide should Ive employed. Tlic .solution
may be standardised against pure ansenicflll) oxide or with a .sodium
thiosulphate solution uhich ha.s been recently standardised against potassium
iodate.

The equation;
lj-f2ce±2r

indicates that the equivalent i.s equal to the atomic weight, or 126,905 p.

Procedure, [’reparation of O.IA'-lodinc. Dissolve 20 g of iodalc-frcc
potassium iodide (e.g.. A.R. I in .'0 -40 cm^ of \sater in a pla.ss-stoppcrcd 1 dm^
graduated flask. Weigh out about 1 2,7 g of A.R. or rcsublimed iodine on a watch
glass on a rough balana' (never on an analytical balance on account of the iodine
vapour), and transfer it In mca.ns of a .small dry funnel into the concentrated
potassium iodide solution. Insert the glass stopper into the flask, tind shake in the
cold until all the iodine has dissolved. Allow the .solution to acquire room
temperature, and make up to the mark with distilled water.

The iodine solution is best prc.scrved in small glass-stoppered bottles. These
should be filled completely and kept in a cool, dark place.

X. 120. .STANDARDISATION OF lODINF, SOLUTIONS. A. With
arscmctlU) oxide. f)iscusst<iii. As already indicated (Section X, 92). A.R.
arsenic(IIl) o.xide which has been dried at 105- 110 C for two hour.s is an excel-
lent primary standard. The reaction between this .substana' and iodine i.s a
reversible one:

MjAsO j -i- 1 , 4- HjO ire 1 ! jA.sQ^ r 21 T 4 21"

und only proceeds quantitatively from left to right if the hydrogen iodide is
removed from the solution a.s fast a-s it is formed, This may be done by the
addition of sodium hydrogencarbonatc; sodium carbonate and sodium
hydroxide cannot be u.scd, .since they react with the iodine, forming iodide,
liypoioditc, and iodate. Actually it has been shown that complete oxidation of the
arsenite occurs when the pH of the solution lies between 4 and 9. the best value
being (y.5. which is very cIo.se to the neutral point. UulTcr solutions arc employed
to maintain the correct pH. A 0.I2A’ solution of .sodium hydrogencarbonatc
saturated with carbon dioxide has a pH of 7; a solution saturated with both
sodium tetraborate and boric acid has a pH of about 6.2. whilst a NtuHPOj--
NalUP 04 solution is almost neutral. .Any of these three buffer solutions is
suitable, but as already stated the first-named is generally employed.

Procedure. Weigh out accurately about 2.5 g of finely powdered A.R.
arscnicfll!) oxide, transfer to a dOO-cm-* beaker, and dissolve it in a concentrated

37.S

TITRIMETRIC ANALYSIS X, 121

solution of sodium hydroxide, prepared from 2 g of iron-free sodium hydroxide
(e.g., A.R.) and 20 cm^ of water. Dilute to about 200 cm^, and neutralise the
solution with M-hydrochloric acid, using a strip of litmus paper as indicator.
When the solution is faintly acid, remove the litmus paper by means of a stirring
rod and carefully rinse both the rod and the paper. Transfer the contents of the
beaker quantitatively to a 500-cm^ graduated flask, add 2 g of pure sodium
hydrogencarbonate, and, when all the, salt has dissolved, dilute to the mark and
shake well. , . ' , ■ ^

Measure out from a burette (this is necessary owing to the poisonous properties
of the solution) 25;0 cm^ of the arsenite solution into a 250-cm^ conical flask, add
25-50 cm^ of water, 5 g of sodium hydrogencarbonate, and 2 cm^ of starch
solution. Swirl the solution carefully until the hydrogencarbonate has dissolved.
Then titrate slowly with the iodine solution, contained in a burette, to the first
blue colour. .

Alternatively, the arsenite solution may be placed in the burette, and titrated
against 25.0 cm^ of the iodine solution contained in a conical flask. When the
solution has a pale yellow colour, add 2 cm^ of starch solution, and continue the
titration slowly until the blue colour is just destroyed.

If it is desired to base the standardisation directly upon arsenic(III) oxide,
proceed as follows. Weigh out accurately about 0.20 g of pure arsenicjlll)’ oxide
into a conical flask, dissolve it in 10 cm^ of M-sodium hydroxide, and add a small
excess of dilute sulphuric acid (say, 12-15 cm^ of N acid). Mix thoroughly and
cautiously. Then add carefully a solution of 2 g of sodium hydrogencarbonate in
50 cm^ of water, followed by 2 cm^ of starch solution. Titrate slowly, with the
iodine solution to the first blue colour. Repeat with two other similar quantities of
the oxide.

B. With standard sodium thiosulphate solution. Sodium thiosulphate
solution, which has been recently standardised, preferably against pure
potassium iodate, is employed. Transfer 25 cm^ of the iodine solution to a 250-
cm^ conical flask, dilute to 100 cm^ and add the standard thiosulphate solution
from a burette until the solution has a pale-yellow colour. Add 2 cm^ of starch
solution, and continue the addition of the thiosulphate solution slowly until the
solution is just colourless.

X, 121. DETERMINATION OF COPPER IN CRYSTALLISED COPPER
SULPHATE. Procedure. Weigh out accurately about 3.0 g of the salt,
dissolve it in water, and make up to 250 cm^ in a graduated flask. Shake well.
Pipette 50.0 cm^ of this solution into a 250-cm^ conical flask, add 1 g potassium
iodide (or 10 cm^ of a 10 per cent solution) ( 1 ), and titrate the liberated iodine with
standard O.lN-sodium thiosulphate (2). Repeat the titration with two other
50-cm^ portions of the copper sulphate solution.

The reaction, written in molecular form, is :

2CuSO,^ + 4JCI = 2CuI + 12 -t- 2 K 2 SO 4
from which it follows that:

2 CUSO 4 = I 2 = 2Na2S203

Note& 1 . If in a similar determination, free mineral acid is present, a few
drops of dilute sodium carbonate solution must be added until a faint permanent

379

X, 122 QUANTITATIVl; INORGANIC ANALYSIS

precipitate remains, and tliis is removed by means of a drop or two ofacctieacid.
Tile potassium iodide is tlicn added and the titration continued. Tor accurate
results, the solution should have a pi I of 4-5.5.

2. After the addition of the potassium iodide solution, run in standard 0.1 A'-
sodium thiosulphate until the brown colouroflhc iodine fades, then add 2crn’ of
starch solution, and continue the addition of the thiosulphate solution until the
blue colour commences to faiie. Then add about 1 g of A.R. potassium or
ammonium thiocyanate, preferably as a 10 per cent aqueous solution: the blue
colour will instantly become more intense. Complete the titration as quickly as
possible. Tlic precipitate possesses a pale flesh colour, and a distinct permanent
end-point is readily obtained.

X, 122. DCTERMINATION OF COPPER IN AN ORfk Disai.csion. Of
the common elements which arc usually associated with copper ores, those that
interfere with the iodornciric determination arc iron, arsenic, and antimony.
Trivalcnl iron is reduced by iodide:

2Fc-'’ -f2r ci2Fe- ^ -t-I;

but by the addition of cxcc'^s of fluoride, the iron(lll) is converted into the
complex [FeF^]^ ". which yields so small a concentration of Fc^ * ions that it has
no oxidising action upon the iodide. .Arsenic and antimony in the trivaleiit form
react with iodine, but in consequence of the oxidising medium usually employed
to bring the sample into solution they will be present in tlie quinqucvalcnl form.
Arscnic(V) and antimoiiy(V( compounds will not oxidise iodide in a solution
having a pi 1 greater than about .s.2. By the use of excess of ammonium hydrogen-
fluoride, .NHiHFj, which acts as a buffer, the pH of the solution can be
maintained above .1.2; under these conditions the reduction of the Cir* ion
proceeds to completion. The concentration of the fluoride should be I.0-!.6Af.

Promiure. The ore (copper pyrites) may be di.ssolvcd in concentrated
nitric acid but it i.s then necessary to evaporate down w iih concentrated sulphuric
acid to remove the nitric acid which would liberate iodine from potassium iodide.
It is therefore preferable to employ perchloric acid which does not give rise to this
problem.

Weigh out accurately about 0.6 g of the dry. finely ground ore into a dry.
narrow-mouth Pyrex flasl:. .Add about 15 cm ' of 72 per cent perchloric acid
(Caution), two small glass K'ads to promote regular ebullition, and insert a short-
necked glass funnel into the neck of the flask. ! Icat the flask geur/y in the fume
cupboard: the acid should reflux down the sides of the flask, but no pronouna'd
white fumes of perchloric acid .should leave the flask. The .sample should di.s.soIvc
in about 5 minutc.s. by which time the condensation ring of acid will have reached
hall-way up the walks of the flask. Remove the burner heiicath the flask and allow
to coo! for 3-4 minutes. Add 50 cm-’ of wmer carefully through the funnel, mix
well, and Imil the solution for 5 minutes. The boiling will remove the free chlorine
formed in the oxidation of the mineral. Cool to room temperature. Add dilute
aqueous ammonia solution (];1) dr<>pwise until the .solution smells slightly of
ammonia. This will precipitate iron hydroxide: an excess of ammonia solution
should be avoided. Add 2.0 g ammonium hydrogenfluoridc NH^HF, and shake
until all the iron hydroxide has dissolved. Now" add 3 g A.R. potassium iodide
di.ssolvcd in 5-10 cm’ water, .and titrate at c>ncc with standard O.LV-sodiuni
thiosulphate, adding 2 cm’ starch solution when the brown colour of the iodine

380

TITRIMETRIC ANALYSIS X, 123/124

decreases in intensity. Continue the addition of the thiosulphate solution until the
blue colour becomes faint. Then add 20 cm^ of 10 per cent aqueous ammonium of
potassium thiocyanate solution, and complete the titration without delay.

X, 123. DETERMINATION OF CHLORATES. Discussion. One pro-
cedure is based upon the reaction between chlorate and iodide in the presence of
concentrated hydrochloric acid :

C 103 --b 6 r-l- 6 H+ =Cl-+3l2-}-3H20
The liberated iodine is titrated with standard sodium thiosulphate solution.

In another method the chlorate is reduced with bromide in the presence of ca.
8 M-hydrochloric acid, and the bromine liberated is determined iodimetrically :

CIO 3 - + 6 Br- + 6 H+ = Cr -bSBri-t-SHjO

Procedure. A. Place 25 cm^ of the chlorate solution (O.liV) in a glass-
stoppered conical flask and add 3 cm^ of concentrated hydrochloric acid
followed by two portions of about 0.3 g each of pure sodium hydrogencarbonate
to remove air. Add immediately about l.Og ofiodate-free potassium iodide and 22
cm^ of concentrated hydrochloric acid. Stopper the flask, shake the contents, and
allow to stand for 5-10 minutes. Titrate the solution with standard O.lN-sbdium
thiosulphate in the usual manner.

B. Place 10.0 cm^ of the chlorate solution in a glass-stoppered flask, add ca.
1.0 g A.R. potassium bromide and 20 cm^ concentrated hydrochloric acid (the
final concentration of acid should be about 8 M). Stopper the flask, shake well,
and allow to stand for 5-10 minutes. Add 100 cm^ of 1 per cent potassium
iodide solution, and titrate the liberated iodine with standard O.lN-sodium
thiosulphate.

X, 124. ANALYSIS OF HYDROGEN PEROXIDE. Discussion. Hydro-
gen peroxide reacts with iodide in acid solution in accordance with the equation ;

H202-t-2H+-b2I- =l2-f2H20

The reaction velocity is comparatively slow, but increases with increasing
concentration of acid. The addition of 3 drops of a neutral 20 per cent ammonium
molybdate solution renders the reaction almost instantaneous, but as it also
accelerates the atmospheric oxidation of the hydriodic acid, the titration is best
conducted in an inert atmosphere (N 2 or CO 2 ).

The iodometric method has the advantage over the permanganate method
(Section X, 95) that it is less affected by stabilisers which are sometimes added.to
commercial hydrogen peroxide solutions. These preservatives are often boric
acid, salicylic acid, and glycerol, and render the results obtained by the
permanganate procedure less accurate.

Procedure. Dilute the hydrogen peroxide solution to ca. 0.3 per cent H 2 O 2 .
Thus, if a ‘20- volume’ hydrogen peroxide is used, transfer 10.0 cm^ by means of a
burette or pipette to a 250 -cm^ graduated flask, and make up to the mark. Shake
well. Remove 25.0 cm^ of this diluted solution, and add it gradually and with
constant stirring to a solution of 1 g of pure potassium iodide in 100 cm^ of JW-
sulphuric acid (1:20) contained in a stoppered bottle. Allow the mixture to stand
for 15 minutes, and titrate the liberated iodine with standard O.lN-sodium
thiosulphate, adding 2 cm^ starch solution when the colour of the iodine has been

nearly discharged. Run a blank determination at the same time. .

381

X, 125 quantitative INORGANIC ANALYSIS

Belter results arc obtained by transferring 25.0 cm^ of the diluted hydrogen
peroxide solution to a conical flask, and adding 100 Af (1 :20) sulphuric acid.
Pass a slow stream of carbon dioxide or nitrogen through the fiask, add 10 cm^ of
10 per cent potassium iodide solution, followed by 3 drops of 3 per cent
ammonium molybdate solution. Titrate the liberated iodine immediately with
standard 0.1 iV-sodium thiosulphate in the usual way.

Note. The above method may also be used for all pcr-salts.

X, 125. DETERMINATIO.N OF THE AVAILABLE CHLORINE IN
BLEACHING POWDER. Dhnifsion. Bleaching powder consists t^sseniially
of a mixture of calcium hypochlorite Ca(OCi), and the basic chloride CaCl,.
Ca{OH),,l LO; some free slaked lime is usually present. Tlic active constituent i*s
the hypochforitc. which is responsible for the bleaching action. Upon treating
bleaching powder with hydrochloric acid, chlorine is lilxeratcd.

OCr + C!”+2H’ Ci, + H,0

The sjvailahle chlorine refers to the chlorine liberated by the action of dilute
acids, and is expressed as the percentage by weight of the bleaching posvdcr. The
bleaching powder of commerce contains Afi-Sf! per cent of available chlorine.

Two methods arc in common use for the determination of the available
chlorine. In the first, the bleaching powder solution or suspension is trcaictl with
an c.xccss of a solution of potassium iodide, and strongly acidified with acetic
acid;

OCT-+2r T2ir liCl- TL.- H.0

The liberated iodine is titrated with standard .sodium thiosulphate solution.
The solution should not be strongly acidified with hydrochloric acid, for the little
calcium chlorate which is usually present, by virtue of the decomposition of the
hypochlorite, will react slowly with the potassium iodide and liberate iodine:

CIO, ■ 4- 61' + 61 r Cr T 31 3 + 3H,0

In the .second method, the bleaching powder solution or .su.spcnsion is titrated
against standard O.LV-sodium arsemte solution; this is bc.st done by adding an
c.xccss of the ai.scnite .solution and then back-titrating with .standard iodine
solution.

Procedure A {irrdomctric method). Weigh out accurately about 5.0 g of the
bleaching powder into a clean glass mortar. Add a little, water, and rub the
mi.xturc to a smooth paste. Adda little more water, triturate with the pestle, allow
the mixture to settle, and pour off the milky liquid into a 500-cm^ graduated flask.
Grind the residue with a little more water, and repeal the operation until the
w'hole of the sample ha.s been transferred to the flask cither in .solution or in a slate
of very fine suspension, and the mortar washed quite clean. The flask is then
filled to the mark with distilled water, well shaken, and 50.0 cm^ of the turbid
liquid immediately withdrawn with a pipette. This is transferred to a 250-cm^
conical flask, 25 ern^ of water added, followed by 2 g of iodatc-frcc potassium
iodide (or 20 cm-’ of a 10 per cent solution) and 10 cm-’ of glacial acetic acid.
Titrate the liberated iodine with standard O.lA'-sodium thiosulphate.

Procedure B (arsenife method). Prepare a bleaching powder .solution
(suspension) as above and transfer 50 cm’ to a 350'Cm’ conical flask. Add from a

382

TITRIMETRIC ANALYSIS X, 126/127

burette 75 cm^ of standard (approximately O.liV) sodium arsenite solution
(Section X, 120), then titrate the excess arsenite with standard (approx. O.IN)
iodine solution.

The concentration of any hypochlorite solution can be determined by either of
the procedures detailed above.

X, 126. DETERMINATION OF ARSENIC(V). The reaction is the reverse
of that employed in the standardisation of iodine with sodium arsenite solution
(Section X, 120);

AS2O5 +4H+ +41- ^AsjOa -I-2I2 + 2H2O

or H3As04 + 2H++2I-^H3As03 + l2 + H20

For good results, the following experimental conditions must be observed: (i) the
hydrochloric acid concentration in the final solution should be at least 4M ; (ii) air
should be displaced from the titration mixture by adding a little solid sodium
hydrogencarbonate; (iii) the solution must be allowed to stand for at least 5
minutes before the liberated iodine is titrated; and (iv) constant stirring is
essential during the titration to prevent decomposition of the thiosulphate in the
strongly acid solution.

Treat the arsenate solution (say, 20.0 cm^ of ca. O.IJV) in a glass-stoppered
conical flask with concentrated hydrochloric acid to give an ca. AM solution in
hydrochloric acid. Displace the air by introducing two 0.4 g portions of pure
sodium hydrogencarbonate into the flask. Add 1.0 g of pure potassium iodide,
replace the stopper, mix the solution, and allow to stand for at least 5 minutes.
Titrate the solution, whilst stirring vigorously, with standard O.lJV-sodium
thiosulphate.

A similar procedure may also be used for the determination of pentavalent
antimony, whilst trivalent antimony may be determined like arsenic(III) by direct
titration with standard iodine solution (Section X, 120 A), but in the antimony
titration it is necessary to include some tartaric acid in the solution; this acts as
complexing agent and prevents precipitation of antimony as hydroxide or as
basic salt in alkaline solution. On the whole, however, the most satisfactory
method for determining antimony is by titration with potassium bromate
(Section X, 139).

X, 127. DETERMINATION OF SULPHUROUS ACID AND OF SUL-
PHITES. Discussion. The iodimetric determination is based upon the
equations:

SO32- +I, + H20 = S04^- +2H+ +2I-

HS03- +I2+H2O = SO^"*- +3H+ +21-

For accurate results, the following experimental conditions must be observed:

(u) the solutions should be very dilute;

(b) the sulphite must be added slowly and with constant stirring to the iodine
solution, and not conversely; and

In determinations of sulphurous acid and sulphites, excess of standard O.IN-
iodine is diluted with several volumes of water, acidified with hydrochloric or
sulphuric acid, and a known volume of the sulphite or sulphurous acid solution is

383

X, 125 QUANTITATJVR INORCMNIC ANAF.VSIS

Better results arc obtained by transferring 25.0 enr' of the diluted hydrogen
peroxide solution ton conical flask, and adding lOOcm’ M (1:20) .sulphuric acid.
Pass a slow stream of carbon dioxide or nitrogen through the flask, add lOcnT* of
10 per cent potassium iodide solution, followed by 3 drops of 3 per cent
ammonium molybdate solution. Titrate the liberated iodine immediately with
standard O.liV-sodium thiosulphate in the usual way.

Note. The above method may also be used for all per-salts.

X, 125. DETERMINATION OE THE AVAILABLE CHLORINE IN
BLEACl ILN G PO^^‘I)F.R. Dheus^on. Bleaching powder consists essentially
of a mixture of calcium hypochlorite Ca(OCl), and the ba.sic chloride CaCl,,.
Ca(OH some free slaked lime is usually present. Tlie active constituent is

the hypochlorite, which is responsible for the bleaching action. Upon treating
bleaching powder sviih hydrochloric acid, chlorine is liberated.

ocr-t cr-^2ir -cu t-HjO

The asnilablc cltlnrinc refers to the chlorine libcr.itcd by the action of dilute
acids, and is expressed as the percentage by weight i>f the bleaching powder. Tlte
bleaching powder of commerce contains 36 per cent nf available chlorine.

Two methods arc in common use for the determination of the available
chlorine. In the lirst, the bleaching powder solution or suspension is treated with
an excc'ss of a solution of potassium iodide, and strongly acidified with aa'tic
acid;

ocr a 2ir 1-11,0

Tile liberated iodine is titrated with standard sodium thiosulphate solution.
The solution should not be strongly acidified with hydrochloric acid, for the little
calcium chlorate which is usually prc.scnt. by virtue of the decomposition of the
hypochlorite, will react slowly with the poiassiurn iodide and lilseratc iodine:

CIO, ■ -f 6r t 6H ‘ Ci •• -i- 31, + 3H,0

In the second method, (he bleaching powder solution or suspension is titrated
against standard 0. LV-sodium arsenite solution; this is best done by adding an
cxcc.ss of (he arsenite solution and then hack-titrating with siand.ard iodine
solution.

Procedure A (iodornctric method) Weigh out .accurately about .NO g of the
blenching powder into .n dean glass mortar. Add a little water, and rub the
mixture to a smooth paste. Add a little more water, triturate with the pestle, allow
the nu’.xturc to settle, and pour off tiic milky liquid into a 500-cm'’ graduated flask.
Grind the residue with a little more water, .and repeat the operation until the
whole of the sample has been iransfctrcd to the flask eitlicr in solution or in a state
of very line suspension, and the mortar washed quite clean. The lla.sk is then
filled to the mark with distilled water, well shaken, and 50.0 enr’ of the turbid
liquid immediately withdrawn with a pipette. This is transferred to a 250-cm’
conical flask. 25 cm’ of water added, followed by 2 g of iodatc-frec potassium
iodide (or 20 cm’ of a 10 per cent solution) and 10 cm’ of glacial acetic acid.
Titrate the liberated iodine with standard 0, 1 A’-sodium thiosulphate.

Procedure IS (arsenite method). Prepare a bleaching powder solution
(suspension) a.s above and transfer 50 cm’ to a 350-cm’ conical flask. Add from a

382

TITRIMETRIC ANALYSIS X, 126/127

burette 75 cm^ of standard (approximately Q.IN) sodium arsenite solution
(Section X, 120), then titrate the excess arsenite with standard (approx. O.IN)
iodine solution.

The concentration of any hypochlorite solution can be determined by either of
the procedures detailed above.

X, 126. DETERMINATION OF ARSENIC(V). The reaction is the reverse
of that employed in the standardisation of iodine with sodium arsenite solution
(Section X, 120); ■

As205 + 4 H^+ 4 I ^As203 + 2I2 + 2H2O

or H3 As 04 + 2H'‘' +2I~ :^H3As03+l2 + H20

For good results, the following experimental conditions must be observed: (i) the
hydrochloric acid concentration in the final solution should be at least 4M; (ii) air
should be displaced from the titration mixture by adding a little solid sodium
hydrogencarbonate; (iii) the solution must be allowed to stand for at least 5
minutes before the liberated iodine is titrated; and (iv) constant stirring is
essential during the titration to prevent decomposition of the thiosulphate in the
strongly acid solution.

Treat the arsenate solution (say, 20.0 cm^ of ca. 0.1 iV) in a glass-stoppered
conical flask with concentrated hydrochloric acid to give an ca. AM solution in
hydrochloric acid. Displace the air by introducing two 0.4 g portions of pure
sodium hydrogencarbonate into the flask. Add 1.0 g of pure potassium iodide,
replace the stopper, mix the solution, and allow to stand for at least 5 minutes.
Titrate the solution, whilst stirring vigorously, with standard O.lN-sodium
thiosulphate.

X, 127. DETERMINATION OF SULPHUROUS ACID AND OF SUL-
PHITES. Discussion. The iodimetric determination is based upon the
equations:

SOj^" +I2 + H2O = SO42- +2H+ +21-

HSO3-+I2 + H2O = SO42-+3H++2I-

For accurate results, the following experimental conditions must be observed:

(a) the solutions should be very dilute;

{b) the sulphite must be added slowly and with constant stirring to the iodine
solution, and not conversely; and
(c) exposure of the sulphite to the air should be minimised.

In determinations of sulphurous acid and sulphites, excess of standard O.liV-
iodine is diluted with several volumes of water, acidified with hydrochloric or
sulphuric acid, and a known volume of the sulphite or sulphurous acid solution is

383

X, 128 QUANTITATIVF. INORGANIC ANALYSIS

added slowly and with consianl stirring from a bvitcUc. with the jet close to the
surface of the liquid. The excess of iodine i.s then titrated with standard O.LV-
sodium thiosulphate. Solid soluble sulphites arc finely powdered and added
directly to the iodine .solution, hrsohiblc sulphites (c.g.. calcium sulphite) react
very slowly, and must be in a very fine state of division.

Procedure. Pipette 2.5.0 cm^ standard fO.IA’) iodine solution into a 350-
enr’ conical flask and add 5 ern^ 2A/-hydrochIoric acid and 150 cm-* distilled
water. Weigh accurately sufiicicnt solid sulphite to react with about 20 cm-* O.IA’-
iodinc soliuion and add this to the content.s of the f1a.sk; swirl the liquid until all
the solid ha.s dissolved and then titrate the c.xce.ss iodine with standard (O.hV)
sodium thiosulphate using starch indicator. If the sulphite is in solution, then a
volume of this equivalent to about 20 cm’ ofO.hV-iodine should be pipetted into
the contents of the fla.sk in place of the weighed amount of solid.

X. 128. DirrivRMlN.VnON or hydrogicn sulphide and

SULPHIDIvS. Discussio)}. The iodimetric method utilises the reversible
reaction

HiS-f I,ee:2ir •f-2r +S

For re;rsonably satisfactory results, the sulphide solution must be dilute
(concentration not greater than 0,04 per cent or 0.02iV), and the sulphide solution
added to excess of acidified O.OLV- or O.l.V -iodine and not convcnscly. Loss of
hydrogen sulpiiide is thus avoided, and side reactions are almo.st entirely
ciiminaicd. (With solutions more concentrated than about n.02A', the
precipitated .sulpluir encloses a portion of the iodine, atid this escapes the
subsequent titration with the .standard sodium thiosulphate. solution.) The e.xccss
of iodine is then titrated with standard thiosulphate solution, using starch as
indicator.

nxcellent results are obtained by the following method, which is of wider
applicability. Wlicn excess of standard sodium arsenite solution is treated with
hydrogen sulphide solution and then acidified with hydrochloricacid.arscnirflll)
sulphide is precipitated:

As-Oj + 311,8 As,S,-f 3H,0

The cxce.ss of arsenicfl II) oxide is determined with 0. 1 A'-iodine and starch.

The procedure is illustrated by detennination of the strength of hydrogen
sulphide water.

Procedure. Prepare a saturated .solution of hydrogen sulphide by bubbling
the gas through distilled water. Place 50.0 cm-* standard 0. 1 A’-sodiiim ar.scnitc in
a 250-cnr’ graduateil flask, add 20cm‘’ of the hydrogeti sulphide water, mix well,
and add sufiicicnt hydrochloric acid to render the solution distinctly acid. A
yellow precipitate of arsenicdlll sulphide is formed, but the liquid itself is
colourless. Make up to the mark with distilled water, and shake thoroughly.
Filter the mixture through a dry filter paper into n dry vessel. Remove 100 env’ of
the filtrate, neutralise it with sodium hydrogcncarbonate, and titrate with
standard 0. 1 A'-iodine to the first blue colour with starch. The quantity of residual
arscnicfll!) oxide is tluis determined, and is deducted from the original 50 cm^
employed.

Note. If certain sulphides are treated with hydrocliloric acid, hydrogen
sulphide is evolved and can be absorbed in an ammoniacal cadmium chloride
solution: upon acidilicalion hydrogen sulpiiide is released.

384

TITRIMETRIC ANALYSIS X, 129/130

Hydrogen sulphide and soluble sulphides can also be determined by oxidation
with potassium iddatein an alkaline medium. Mix 10.0 cm^ of the sulphide solution
containing about 2.5 mg sulphide with 15.0 cm^ O.lJV-potassium iodate (Section
X, 132) and 10 cm^ of lOM-sodium hydroxide. Boil gently for 10 minutes, cool,
add 5 cm^ of 5 per cent potassium iodide solution and 20 cm^ of 4M-sulphuric
acid. Titrate the liberated iodine, which is equivalent to the unused iodate, with
standard 0. 1 /V-sodium thiosulphate in the usual manner.

X, 129. DETERMINATION OF HEXACYANOFERRATES(III). Discus-
sion. The reaction between hexacyanoferrates(III) (ferricyanides) and soluble
iodides is a reversible one:

2[Fe(CN)e]^- +21- -2[Fe(CN)6f - +I 2

In strongly acid solution the reaction proceeds from left to right, but is reversed in
almost neutral solution. Oxidation also proceeds quantitatively in a slightly acid
medium in the presence of a zinc salt. The very sparingly soluble potassium zinc
hexacyanoferrate(II) is formed, and the hexacyanoferrate(II) ions are removed
from the sphere of action:

2[Fe(CN)6]^- +2K+ +3Zn^^ = K2Zn3[Fe(CN)6]2

The procedure may be used to determine the purity of potassium
hexacyanoferrate(III).

Procedure. Weigh out accurately about 10 g of the salt and dissolve it in
250 cm^ of water in a graduated flask. Pipette 25 cm^ of this solution into a 250-
cm^ conical flask, add about 20 cm^ of 10 per cent potassium iodide solution, 2
cm^ of M-sulphuric acid, and 15 cm^ of a solution containing 2.0 g crystallised
zinc sulphate. Titrate the liberated iodine immediately with standard 0.1/V-
sodium thiosulphate and starch; add the starch solution (2 cm^) after the colour
has faded to a pale yellow. The titration is complete when the blue colour has just
disappeared. When great accuracy is required, the process should be conducted
in an atmosphere of carbon dioxide.

X,130, STANDARDISATION OF AN ACID. When an iodate is allowed to
react with iodide ions in solutions of moderate acidity, free iodine is liberated and
it is apparent from the equation for the reaction (Section X, 115) that the amount
of iodine liberated is equivalent to the acid content of the solution provided that
excess of iodate and iodide are present.

Procedure. Pipette 25 cm^ of the acid solution to be standardised into a
250-cm^ conical flask, add 1.0-1.5 g potassium iodide crystals or 10 cm^ of a 10
per cent solution of potassium iodide, followed by 25 cm^ of 0.5 per cent
potassium iodate solution. Titrate the liberated iodine with standard sodium
thiosulphate solution.

When dealing with a solution containing a weak acid, the rate of reaction is
rather slow, and it is then preferable to add 50 cm^ of standard sodium
thiosulphate solution after adding the potassium iodate solution. The
thiosulphate removes the iodine as it is liberated, and the speed of reaction is
increased. Allow the solution to stand for ten minutes after adding the
thiosulphate and then back titrate the excess thiosulphate with a standard iodine
solution.

385

X, 131 QUANTITATIVE INORGANIC ANALYSIS

Oxidations with potassium iodate

X, 131. GENERAL DISCUSSION. Potassium iodate is a powerful oxidising
agent, but tiic course of the reaction is governed by the conditions under which it
is cmpioyed. Tiie reaction between potassium iodate and reducing agents such as
iodide ion or arsenictllt) oxide in .solutions of moderate acidity (0,1-2.0.W
hydrochloric acid) stops at the stage when the iodate is reduced to iodine:

io.r+5r +f.ir 311,0

2IOj- + ^I, + 5 M,.As04 + ILO.

As already indicated (Section X, 1 15). the first of these reactions is very useful for
the generation of known amounts of iodine, and it also serves as the basis of a
method for standardising solutions of acids (Section X, 130),

With a more powerful rcductant. e.g.. litaniumdll) chloride, the iodate is
reduced to iodide;

lOj- +fiTi^* -) 6H’ ^ 1“ +6Ti*’ -f.3H,0

In more strongly acid solutions (3-6.1/-hydrochloric acid) reduction occurs to
iodine rnonochloridc, and it is under these conditions (due to Andrews) that it is
most widely used (Refs 14, 15).

1 Oj a- 611 ‘ -I- Cl ' *»- 4e re: f Cl 4-311,0

In hydrochloric acid .solution, iodine rnonochloridc forms a stable compIc.x ion
with chloride ion:

ICl-fCT- relClC

The overall half-eel! reaction may therefore be written as:

10, ~ +6ir 4-20* 4-4 c=^K:1,' -*-311:0;

the reduction potential is 1.23 volts, hence under these conditions potassium
iodate acts as a very powerful o.xidising agent, rurlhcrmorc. urJcr ihesc
pariiculiir coiulitiaita the equivalent of potassium iodate is one-fourth of a mole
RIO, ,'4. and a 0.1 A' solution will contain K10r'(4 -x 10), or 214.00/40 = 5.3500 g
dm ■ \ This is by contrast with the situation where reduction to iodine occurs (ic.,
conditions of mild acidity), wlien the equivalent is one-sixth of a mole and a O.LV
solution contains 214.00/6 = 3.5667gdm'-’.

Oxidation hy iodate ion. in a strong liydrochloric acid medium proceeds
through several" stages:

10,--f6Fr-f6e=±l' +311,0

10,-+5I-+6H" - 3l2 + 3M:0

10,-+2l3 + 6H" =5r +311,0

In the initial stages of the reaction free iodine is liberated; as more titrant is added,
oxidation proceeds to iodine rnonochloridc. and the dark colour of the solution
gradually disappears. The overall reaction may be written as;

lO,' +6H" +4e:?i:r +3H,0

The reaction has been used for tlic determination of many reducing agents: the
optimum acidity for reasonably rapid reaction varies from one rcductant to

386

TITRIMETRIC ANALYSIS X, 132/133

another within the range 2 . 5 - 9 M-hydrochloric acid; in many cases the
concentration of acid is not critical, but for Sb(III) it is 2.5-3.5M.

Under these conditions starch cannot be used as indicator because the
characteristic blue colour of the starch-iodine complex is not formed at high
concentrations. of acid. In, the original procedure, a few cm^ of an immiscible
solvent (carbon tetrachloride or chloroform) were added to the, solution being
titrated contained in a glass-stoppered bottle or conical flask. The end-point.is
marked by the disappearance of the last trace of violet colour, due to iodine, from
the solvent; iodine monochloride is not extracted and imparts a pale yellowish
colour to the aqueous phase. The extraction end-point is very sharp. The main
disadvantage is the inconvenience of vigorous shaking with the extraction solvent

in a stoppered vessel after each addition of the reagent near the end-point

The immiscible solvent ihay be replaced.by certain.dyes, e.g.. Amaranth (C.I.
161 85),. colour change red to colourless; Xylidine Ponceau (C.I. 16255), colour
change orange to colourless; Naphthalene Black 12B (C.I. 20470), colour change
green to faint pink; the first two of these are generally preferred. The indicators
are used as 0.2 per cent aqueous solutions and about 0.5 cm^ per titration is added
near the end-point. The dyes are destroyed by the first excess of iodate, and hence
the indicator action is irreversible. The indicator blank is equivalent to 0.05 cm^
of O.lN-potassium iodate per 1.0 cm^ of indicator solution, and is therefore
virtually negligible. , i

p-Ethoxychrysoidine is a moderately satisfactory reversible indicator. It is used
as a 0.1 per cent solution in ethanol (about 12 drops per titration), and the colour
change is from red to orange; the colour is red-purple just before the end-point.
The indicator is added after the colour of the iodine commences to fade. A blank
determination should be made for each new batch of indicator.

X, 132. PREPARATION OF 0.025M-POTASSIUM IODATE. Dry some
A.R. potassium iodate at 120 °C for 1 hour and allow it to cool in a covered vessel
in a desiccator. Weigh out exactly 5.350 g of the finely powdered potassium iodate
on a watch glass, and transfer it by means of a clean camel-hair brush directly into
a dry 1-dm^ graduated flask. Add about 400-500 cm^ of water, and gently rotate
the flask until the salt is completely dissolved. Make up to the mark with distilled
water. Shake well. The solution will keep indefinitely.

It must be emphasised again that the solution is O.llV only for the reaction:

I 03 -+ 6 H++Cr +4e^ICl-t-3H20 ,

and when used in solutions of moderate acidity leading to the liberation of free
iodine, the O.IAT solution requires 3.5667 g KIO 3 Per litre; the method of
preparation will be as described above with suitable adjustment of the weight of
salt taken.

X, 133. DETERMINATION OF ARSENIC OR OF ANTIMONY. Discus-
sion. The determination of arsenic in arsenic(III) compounds is based upon the
following reaction:

I03-q-2H3As03-}-2H+-t-Cl- =IC1+2H3As04-1-H20
A similar reaction occurs with antimony(III) compounds. The determination
of antimony(III) m the presence of tartrate is not very satisfactory with an
immiscible solvent to assist in indicating the end-point; Amaranth, however

fnVPC PYr'plipnf r<acnlfc ’ ’

387

X, 134 QUANTITATIVE INORGANIC ANALYSIS

I 03 -+ 2 (SbCl 4 ]--f-GH^ +5Cr - ICl + 2[SbCy'+3H,0

To assay a sample of arscnic(lll) oxitic the following procedure may be used.

Procedure. Weigh out accurately about 1 . 1 g of the oxide sample, dissolve
in a small quantity of warm 10 per cent sodium hydroxide solution, and make up
to 250 cm-' in a graduated flask. Usca hureltc to measure 25.0cm^ of this solution
into a stoppered reagent bottle of about 250 cm-* capacity.* add 25 cm^ water. 60
cm*’ concentrated hsdrochloric acid and about 5 cm-' carbon tetrachloride or
chloroform. Coo! to roont temperature. Run in the standard O.lN-potassium
iodatc from a burette until the solution, which at first is strongly coloured wiih
iodine, becomes pale brown. 'Hie bottle is then stoppered and vigorously shaken,
and the organic solvent layer acquires the purple colour due to iodine. Continue
to add .srnall volumes of the iodatc solution, shaking vigorously after each
addition, until the organic layer is only very faintly violet. Continue the addition
dropwisc. with shaking after each drop, until the solvent loses the last trace of
violet and has only a very pale-yellow colour (due to iodine chloride). Tlie end-
point is very sharp and. .ificr a little experience, is rarely overshot. If this should
occur, a small volume of the oxide .solution is added from a graduated pipette,
and the end-point re-determined. Allow to stand for ten minutes and observe
whether the organic layer shows any purple colour; the absence of colour
confirms that the titration is complete.

The acidity of the mixttirc at the end of the titration should be not less than 351
and not more than 5.\/: if the acidity is too high the reaction takes place slowly.

X,134. DETERMINATION OF MKRCURV. Diratssinii. Tiic mercury is
precipitated as mcrcuryll) chloride and the latter is rctictcd with standard
potassium iodatc solution:

103 '-f- 2 HgXI; + f.ir -f 1.3CI - IC:iT4[HgCl4)=’'-f3HjO.

Thus KIOj s?4Hg-- 2Mg2CI-.

To determine the purity of a s.ample of a mcrcuryfll) salt the following
procedure in which the compound is reduced with phosphorous acid may be
used ; to assay a sample of a mcrcuryll) salt, the reduction with phosphorous acid
is omitted.

Procedure. Weigh out acc'- ■^.itcly about 2.5 g of finely powdered
mcrcury(ll) chloride, and dissolve it in i(Xl cm’ of water in a graduated flask.
Shake well. Transfer 2.5.0 cm ’ of the solution to a conical flask, add 25 cm’ water,
2 cm’ A'-hydrochloric acid, and excess of 50 per cent phosphorous acid solution.
Stir thoroughly and allow to stand for 12 hours or more. Filter the precipitated
mcrcury(I) chloride through a quantitative filter paper or through a Gooch
crucible with aslx'stos, and wash the precipitate moderately with cold water.

• A 250-cin’ pradii.'itcd fl.nst. with .i stiorl neck and .n wcll-fmins,’ ground pla's stopper ma> also he
used. The colour of the orp.inic l;u cr is readily seen by inwriinp the H.asl, so that the layer of solvent
indicator collects m the neck.

Alternatively, in this and all subsciincnt iiirntions with O.I A'-pol.issiiin! iodatc. a 250- or ?50<ni’
conical flask m.ny be used and the carbon tetrachloride or cliloroform indicator replaced by 0.5 cm’
Amaranth or .Vylidinc I’onccau indie.alor. which is added after most of the iodine colour has
disappeared from the reaction mixture (sec Section X. t.tl).

388

. TITRIMETRIC ANALYSIS X, 135

Transfer the precipitate with the filter paper or asbestos quantitatively to a 250-
cm^ reagent bottle, add 30 cm^ concentrated hydrochloric acid, 20 cm^, water,
and 5 cm^ carbon tetrachloride or chloroform. Titrate the mixture with standard
O.lN-potassium iodate in the usual manner (Section X, 133).

2HgCl2 + H 3 P 03 -f-H 20 = Hg2Cl2-t-2HCl + H3P04

Many other metallic ions which are capable of undergoing oxidation by
potassium iodate can also be determined. Thus for example copper(II) compounds
can be analysed by precipitation of copper(I) thiocyanate which is titrated with
potassium iodate:

7 IO 3 ~ 4 - 4CuSCN '+ 1 8H + -h 7C1 " = . 7IC1 -t- 4Cu^ + -I- 4 HSO 4 " -f 4HCN

-(- 5 H 2 O.

As a typical example, 0.8 g of copper(II) sulphate CuS 04 , 5 H 20 is dissolved in
water, 5 cm^ of 0.5M-sulphuric acid added, and the solution made up to 250 cm^
in a graduated flask. 25.0 cm^ of the resulting solution are pipetted into a 250-cm^
conical flask, 10-15 cm^ of freshly prepared sulphurous aeid solution added, and
then after heating to boiling, 10 per cent ammonium thiocyanate solution is
added slowly from a burette with constant stirring until there is no further change
in colour, and then 4 cm^ of reagent is added in excess. After allowing the
precipitate to settle for 10-15 minutes, it is filtered through a Gooch crucible
containing asbestos and then washed with cold 1 per cent ammonium sulphate
solution until free from thiocyanate. It is then transferred quantitatively into the
vessel in which the titration is to be performed, and after adding 30 cm^ of
concentrated hydrochloric acid, followed by 20 cm^.of water, the titration is
carried out in the usual manner with either an organic solvent present, or an
internal indicator is added as the end-point is approached.

Thallium(I) salts are oxidised in accordance with the equation :

IO3 - -f 2T1+ + 6 H+ 4- Cl - = ICl + 2TP + + 3H2O

so that KIO 3 = 2T1.

The solution should contain 0.25-0.30 g Tl"^ in 20 cm^ plus 60 cm^ of
concentrated hydrochloric acid and is titrated as usual with 0.025M KIO 3
solution.

Tin(n) salts are likewise oxidised in accordance with the equation

I 03 -'-h 2 Sn 2 + 4 - 6 H+ 4 -Cr = ICl4-2Sn'‘+ 4 - 3 H 2 O.

so that KIO3 = 2Sn.

If the bulk of the iodate solution is added rapidly, atmospheric oxidation does not
present a serious problem, but the method cannot be used in the presence of salts
of antimony(III), copper(I) or iron(II). The solution which should contain for
example 0.15 g SnCl2,2H20 in 25 cm^ is treated with 30 cm^ of concentrated
hydrochloric acid and 20 cm^ of water and is then titrated in the usual manner
with standard potassium iodate solution.

X, 135. DETERMINATION OF HYDRAZINE. Discussion. Hydrazine
reacts with potassium iodate under the usual Andrews conditions thus;

I03-4-N2H4+2H+4-Cr =ia4-N2-b3H20

389

X, 136/137 QUANTITATlvn INORGANIC ANALYSIS

Thus KlOj ~ NjHj

To determine llic content of hydra/inium sulphate, use the

following method.

Procedure. Weigh out accurately 0.08-0. 1 g of hvdrazinium sulpliatc into a
250-cm^ reagent bottle, add a mixture of 30 cm’ of concentrated hydrochloric
acid, 20 cm' of water, ;tnd 5 cm' of chloroform or carlmn tetrachloride. Run in the
.standard 0.025A/-potii.ssium iodale slowly from a burette, with shaking the
stoppered bottle between the additions, until the organic layer is just decolorised.

X. 136. DETERMl.NATION OF VANADATFiJ. Dheumou. Vanadates
arc reduced by iodides in strongly acid (hydrochloric) solution in an atmosphere
ofc.arbon dioxide to the quadrivalent condition:

2 V Oj ' ■ -t- 21" -M 21 r 2VO' * a- 1 : 4- 6H ; O
Tltc liberated iodine and the c.vccss of iodide is determined by titration with
standard pota.ssium iodatc solution; the hydrochloric acid concentration ntu-sl
not be allowed to fitll below 7.\f in order to prevent re-oxldation of the vanadium
compound by iodine chloride.

21, -MO.,' 4 61 r 4 5rr - sich 3n.o

21- 4-10>”4-6H’ -t-3C:r 3lCl4-.3H,0

The tola! result of the reaction is:

4VO,’ ■ 4-41" 4- lO.,- 4 5Cr 4'.'«0H ' r-. 4VO'' 4-5IC14- 15H,0

This method is applicable in the presence of ar.scn:tte. phosphate, or tronfllll,
and also in the presence of tungrstic acid, which may be held in solution by adding
phosphoric acid.

Procedure. Place 25.0 cm' of the solution containing 0.05-0.10 g of
vanadium (as vanadate) in a 250-cm' glass-.stoppcrcd reagent bottle, and p.as.s a
rapid current of carbon dioxide for 2 3 minutc.s into the bottle, but not through
the solution. Then add suflicient concentrated hydrochloric acid through a funnel
to make the solution 6-8, \/ during the titration. Introduce a known volume
(excess) of tipproximaicly O.OS.M-poIassium iodide, which ha.s Ixren titrated
against the standard iodatc solution. Mix the contents of the hottlc. allow to
stand for 1-2 minutes, add 5 cm' of carbon tetrachloride, and then titrate as
rapidly as possible with standard U.()25.M-potassium iodatc until no more iodine
colour can be detected in the organic Liver. .Add concentrated hydrochloric acid
as needed during the titration so that the concentration does not fall below l.M.

Oxidation,s with potassium bromate

X, 137. GFNLRAL DISCUSSION. Potassium bromate is a powerful
oxidising agent which is reduced smoothly to bromide:

BrOj " 4- 61 1 ’ 4- 6t' “ Br ■ + 3H,0

The eqiiivnlent is therefore i mole (KBrOjTd, or 167.00,6, or 27.833, and n O.LV
solution contains 2.7833 g potassium bromate per dm'. .At the end oft he titration
free bromine appears;

BrO^-TSBr' 4-6H' =3Br,+3H,0

390

TITRIMETRIC ANALYSIS X, 137

The presence of free bromine, and consequently the end-point, can be detected by
its yellow colour, but it is better to use indicators such as methyl orange, methyl
red. Naphthalene Black 12B, Xylidine Ponceau, and Fuchsine. These indicators
have their usual colour in acid solution, but are destroyed by the first excess of
bromine. . With all irreversible oxidation indicators the destruction of the
indicator is often premature to a slight extent: a little additional indicator is
usually required near the end-point. The quantity of bromate solution consumed
by the indicator is exceedingly small, and the ‘blank’ can be neglected for 0.1 JV
solutions. Direct titrations with bromate solution in the presence of irreversible
dyestuff indicators are usually made in hydrochloric acid solution, the
concentration of which should be at least 1.5-2M. At the end of the titration some
chlorine may appear by virtue of the reaction: •

10Cr+2BrO3--)-12H-" = 5Cl2-t-Br2-h6H20

this immediately bleaches the indicator. .

The titrations should be carried out slowly so that the indicator change, which
is a time reaction, may be readily detected. If the determinations are to be
executed rapidly, the volume of the bromate solution to be used must be known
approximately, since ordinarily with irreversible dyestuff indicators there is no
simple way of ascertaining when the end-point is close at hand. With the highly
coloured indicators (Xylidine Ponceau, Fuchsine, or Naphthalene Black .12B),
the colour fades as the end-point is approached (owing to local excess of bromate)
and another drop of indicator can be added. At the end-point the indicator is
irreversibly destroyed and the solution becomes colourless or almost so. If the
fading of the indicator is confused with the equivalence point, another drop of
the indicator may be added. If the indicator has faded, the additional.drop will
colour the solution; if the end-point has been reached, the additional drop of
indicator will be destroyed by the slight excess of bromate present in the solution.
The introduction of reversible redox indicators for the determination of
trivalent arsenic and trivalent antimony has considerably simplified the
procedure; those at present available include 1-naphthoflavone, and p-
ethoxychrysoidine. The addition of a little tartaric acid or potassium sodium
tartrate is recommended when antimony(III) is titrated with bromate in the
presence of the reversible indicators; this will prevent hydrolysis at the lower acid
concentrations. The end-point may be determined with high precision by
potentiometric titration (see Chapter XIV).

Examples of determinations utilising direct titration with bromate solutions
are expressed in the following equations :

(HCl)

BrOs - -h 3H3ASO3 — ► Br- + 3H3 ASO4
(HCl)

2Br03--t-3N2H4 ^ 2Br- +3N2-1-6H20

„ (HCl)

Br03-+NH20H — > Br' -1-NO3- -I-H+ -fH20
Br03- -l-6[Fe(CN)6]^- +6H+ — ^ Br” -!-6[Fe(CN)6]3- -f 3H2O

Various substances cannot be oxidised directly with potassium bromate but
react quantitatively with an excess of bromine. Acid solutions of bromine of

391

X, 138/139 QUANTITATIVE INORGANIC ANALYSIS

exactly known concentration arc readily obtainable from a standard potassium
bromate solution by adding acid and an excess of bromide:

BrO,' + 5Br" -f-6ir "=3Br2+3H}0

In this reaction 1 mole of bromate yields six atoms of bromine, hence the
equivalent is KBrOj/6, identical with that of potassium bromate alone. Bromine
i.s verv volatile, and hence .such operations should be conducted at as low a
temperature as po.ssiblc and in conical fla.sks fitted with ground-glass stoppers.
The e.xcess of bromine may be determined iodomelrically by the addition of
excess of potassium iodide and titration of the liberated iodine with standard
thiosulphate solution:

2r +Br; l,H-3Br

Potassium bromate is readily available in a high state of purity; the A.R.
product has an assay value of at iea.st 99.9 per cent. The .substance can be dried at
120-150 X. is anhydrous, and the aqueous solution keeps indetinitcly. It can
therefore i>e employed as a primary standard. Its only disadvantage is that the
equivalent is comparatively small

X, 138. PREPARATION OP O.f N-POTASSU)M BROMATIi. Dry .some
finely powdered A.R. potassium bromate for I 2 hours at 1 20 C. and allow to
cool in a closed vessel in a desiccator. Weigh out accurately 2.7S3 g of the pure
potassium bromate. and dissolve tt in 1 dm ' of water in a graduated flask.

X. 139. DlsTERMlNATlON OF ANTIMONY OR OF AR.SENIC D'mis-
sioii. The antimony or the arsenic must be present in the irivalent condition.
The reaction of trivalent arsenic or antimony with pottissiuin bromate may be
written;

2KBr0.,-f-3M;0,-t 2HCI ’KCl -f dMjO. i ^UBr (M .AsorSb)

The presence of tin and of considerable quantities of iron and copper interfere
with the determinations.

To determine the purity of a .sample of at.senic(Hl) oxide follow the general
procedure outlined in Section X. 133 but when the 25-cm^ .sample of .solution is
being prepared for titration, add 25 cm' water. 15 cm' of concentrated
hydrochloric acid and then two drops of indicator solution (Xylidine Ponceau or
Naphthalene Black I2B; .see Section X, 131), Titrate slowly with the standard
0. 1 ,V-poias.sium bromate with constant swirling of the solution. As the end-point
approaches, add the bromate solution dropwise with inteivals of 2-3 seconds
between the drops until the solution is colourie.ss or very pale yellow. Iftiiecolour
of the indicator fade.s. add another drop of indicator solution. (The immediate
discharge of ihc colour iiuJicalcs that the equivalence point has been passed and
the titration is of little value.)

As an alternative, a reversible indicator may be employed, either (n) I-
naphthoflavone {0.5"„ .solution in ethanol, wliicii gives an orange-coloured
solution at the end-point), or (h) p-cthoxychrysoidine (0.1% aqueous solution,
colour change pink to pale yellow). Under these conditions, the measured 25-cm'
portion of the arsenic solution is treated with 10 cm' of 10 per cent potassium
bromide solution, 6 cm' of concentrated hydrochloric acid. 1 0 cm' of water and
either 0.5 cm' of indicator (a) or two drop.s'of indicator (b).

392

TITRIMETRIC ANALYSIS X, 140

X, 140. DETERMINATION OF METALS BY MEANS OF 8-
HYDROXYQUINOLINE (‘OXINE’). Discussion. Various metals (e.g.,
aluminium, iron, copper, zinc, cadmium, nickel, cobalt, manganese, and
magnesium) under specified conditions of pH yield well-defined crystalline
precipitates with 8-hydroxyquinoline. These precipitates have the general
formula M(C9H60N)„, where n is the valency of the metal M (see, however.
Section XI, IIC). Upon treatment of the oxinates with dilute hydrochloric acid,
the oxine is liberated. Oxine reacts with 4 equivalents of bromine to give 5,7-
dibromo-8-hydroxyquinoline :

C 9 H 70 N 4 - 2 Br 2 = CgHsONBra-blH^ -t-2Br-

Hence 1 mole of the oxinate of a divalent metal requires 8 equivalents of bromine,
whilst that of a trivalent metal requires 12 equivalents. The bromine is derived by
the addition of standard O.lN-potassium bromate and excess of potassium
bromide to the acid solution.

Br03“-h5Br--f6H+ = SBr^-fSHjO

Full details are given for the determination of aluminium by this method.
Many other metals may be determined by this same procedure, but in many cases
complexometric titration olfers a simpler method of determination. In cases
where the oxine method offers advantages, the experimental procedure may be
readily adapted from the details given for aluminium.

Determination of aluminium. Prepare a 2 per cent solution of A.R.
8-hydroxyquinoline (see Section XI, IIC) in 2M acetic acid; add ammonia solu-
tion until a slight precipitate persists, then redissolve it by warming the solution.

Transfer 25 cm^ of the solution to be analysed, containing about 0.02 g of
aluminium, to a conical flask, add 125 cm^ of water and warm to 50-60 °C. Then
add a 20 per cent excess of the oxine solution (1 cm^ will precipitate 0.001 g of Al),
when the complex A1(C9H60N)3 will be formed. Complete the precipitation by
the addition of a solution of 4.0 g of ammonium acetate in the minimum quantity
of water, stir the mixture, and allow to cool. Filter the granular precipitate
through a sintered glass crucible of porosity No. 4 (or through a porcelain
filtering crucible), and wash with warm water (1). Dissolve the complex in warm
concentrated hydrochloric acid, collect the solution in a 250-cm^ reagent bottle,
add a few drops of indicator (0.1 per cent solution of the sodium salt of methyl red
or 0.1 per cent methyl orange solution), and 0.5-1 g of pure potassium' bromide.
Titrate slowly with standard O.UV (i.e., M/60) potasium bromate until the colour
becomes pure yellow (with either indicator). The exact end-point is not easy to
detect, and the best procedure is to add an excess of potassium bromate solution,
i.e., a further 2 cm^ beyond the estimated end-point, so. that the solution now
contains free bromine. Dilute the solution considerably with 2M-hydroehloric
acid (to prevent the precipitation of 5,7-dibromo-8-hydroxyquinoline during the
titration), then add (after 5 minutes) 10 cm^ of 10 per cent potassium iodide
solution, and titrate the liberated iodine with standard O.UV-sodium
thiosulphate, using starch as indicator (2).

From the above discussion, it is evident that Al = 12Br, i.e., to 12 dm^ of IJV-
bromate (or 1 V-thiosulphate).

Notes. 1. This will remove the excess of oxine. Complications due to
adsorption of iodine will thus be avoided.

2. A brown additive compound of iodine with the dibromo compound may

393

X, I4I/J42 QUANTITATJVF. INORGANIC ANALYSIS

separate during the titration; tltis compound usually dissolves during the
subsequent titration with thiosulphate, yielding a yellow solution so that (he end-
point with starch may be found tn the usual manner. Occasionally, the dark-
coloured compound, which contains nd.sorbcd iodine, may not dissolve readily
and thus iniroducc.s an uncertainty in (he end-point; this difficulty may be
avoided by adding 10cm’ ofe-arbon disulphide before introducing the potassium
iodide solution.

X. 141. DE'l'KRMlNATION OF HYDUOXY1.A.MINE The method based
upon the reduction ofirontHl) solutions in the presence ofsuiphuric acid, boiling
and subsequent titration in the cold witlr standard O.hV-potassium
permanganate frequently yields high results unless the experimental conditions
arc closely controlled:

2NH;01H 4Fe’* - N,0-s 4lV' 4-41!" -t H.O

Better results are obtained In oxidation with potassium bromatc in the presence
of hydrochloric acid :

NII;OH-f BrO., ■ N'Or +Rf-i-ir T 11-0

The hydro.xylarninc .solution is treated with a measured volume of 0.1 A' (i.c.,
potass'ium bromatc so as to give 10-30 env’ excess, followed by 40 cm-’ of
53/-hydrocliloric acid. After 15 minutes the e.xccss of bromatc is determined by
the addition of potassium iodide solution and titration with standard O.IX-
sociium thiosulphate (compare Section X, 140).

The Reduction of Higher Oxidation States

X, 142. GENICRAl. DISCUSSION. It has already Isccn indicated that before
titration with an oxidising agent can be carried out. it may in some cases be
necessary to reduce the compound supplied to a lower state ofo.xidation. Such a
situation is frequently encountered with the determination of iron; ironflUl
compounds must be reduced to iron(II) before titration with potassium
permanganate or potassium dichromatc can he performed. It is possible to carry
out sucli determinations directly as a raliiciinu-iric titration by the use of
solutions of powerful reducing agents such as chromiumfll) chloride,
titaniumlllltchlorideorvanadiumflD.sulphaie. but the problcm-s associated with
the preparation, storage and handling of these reagents have militated against
their widc,sprcad use. Titaniumf! 1 {) sufphatc has found application in the analysis
of certain types of organic compounds (Ref, 3), but is of limited application in the
inorganic field. An apparatus .suitable for the preparation, storage and
manipulation of chromiumfll) and vanadiumlll) solutions is described in
Reference 16; with both these reagents it i.s necessary (and isai.so adxisable with
Ti(lII) solutions), to carry out titrations in an atmosphere of hydrogen, nitrogen
or carbon dioxide, and in view of the instability of nio.st indicatonsin the presence
of these powerful reducing agents, it is frequently nccessarv to determine the end-
point potciuiomctrically.

The most important method for reduction of compounds to an oxidation state
suitable for titration with one of the common oxidising titranls is based upon the
use of metal amalgams, but there are various other methods which can be used,
and these will be discussed in tiic following Sections.

394

TITRIMETRIC ANALYSIS X, 143

X,143. REDUCTION WITH AMALGAMATED ZINC; THE JONES
REDUCTOR. Amalgamated zinc is an excellent reducing agent for many
metallic ions. Zinc reacts rather slowly with acids, but upon treatment with a
dilute solution of a mercury(II) salt, the metal is covered with a thin layer of
mercury; the amalgamated metal reacts quite readily. Reduction with
amalgamated zinc is usually carried out in the ‘reductor’, due to C. Jones. This
consists of a column of amalgamated zinc contained in a long glass tube provided
with a stopcock, through which the solution to be reduced may be drawn. A large
surface is exposed, and consequently such a zinc column is much more efficient
than pieces of zinc placed in the solution.

A suitable form of the Jones reductor, with approximate dimensions, is shown
in Fig. X, 18. A perforated porcelain plate, covered with purified asbestos or glass
wool, supports the zinc column. The tube below the tap passes through a tightly

fitting one-holed rubber stopper into a 750-cm^
filter flask. It is advisable to connect another filter
flask in series with the water-pump, so that if any
water ‘sucks back’ it will not spoil the deter-
mination. The amalgamated zinc is prepared as
follows. About 300 g of A.R. granulated zinc (or
zinc shavings, or pure 20-30-mesh zinc) are covered
with 2 per cent mercury(II) chloride solution in a
beaker. The mixture is stirred for 5-10 minutes,
then the solution is decanted from the zinc, which is
washed three times with water by decantation. The
resulting amalgamated zinc should have a bright
silvery lustre. The porcelain plate is placed in
position, covered with a layer of purified asbestos or
glass wool and then the amalgamated zinc added :
the latter should reach to the shoulder of the tube.
The zinc is washed with distilled water (500 cm^),
using gentle suction. If the reductor is not to be used
immediately, it must be left full of water in order to
prevent the formation of basic salts by atmospheric
oxidation, which impair the reducing surface. If the
moist amalgam is exposed to the moisture of the
atmosphere, hydrogen peroxide may be generated :

Zn -h O3 -1- 2H2O = Zn(OH)2 + H2O2

but no hydrogen peroxide is formed if acid is present.

To use the reductor for the reduction of iron(in), proceed as follows. The zinc
is activated by filling the cup (which holds about 50 cm^) with M (ca. 5 per cent)
sulphuric acid, the tap being closed. The flask is connected to a filter pump, the
tap opened, and the acid slowly drawn through the column until it has fallen to
just above the level of the zinc; the tap is then closed and the process repeated
twice. The tap is shut, the flask detached, cleaned, and replaced. The reductor is
now ready for use. It is important to note that during use the level of the liquid
should always be just above the top of the zinc column. The solution to be
reduced should have a volume of 100-150 cm^ contain not more than 0.25 g of
iron, and be about M in sulphuric acid. The cold iron solution is passed through
the reductor, using gentle suction, at a rate not exceeding 75-100 cm^ per minute

45 mm

395

TITRIMETRIC ANALYSIS X, 144

alum and 150 cm^ of concentrated sulphuric acid per litre;. approximately 03N
with respect to iron) contained in the filter flask. The iron(II) formed is then
titrated with a standard solution of a suitable oxidising agent. Titanium and
chromium are completely oxidised and produce an equivalent amount of iron(II)
sulphate; molybdenum is reoxidised ;to the quinquevalent (red) stage, which i§
fairly stable in air, and complete oxidation is effected by the permanganate, but
the net result is the same, viz., Mo(III) Mo(Vl); vanadium is re-oxidised to the
quadrivalent, condition, which is stable in air, and the final oxidation is
completed by slow titration with potassium permanganate solution or with
cerium(IV) sulphate solution. ■ ' .

X, 144. REDUCTION WITH LIQUID AMALGAMS. Makazono in-
troduced liquid zinc amalgam as a reducing agent and. subsequent Japanese
workers have used liquid amalgams of cadmium, bismuth, and lead. The
advantages claimed for liquid amalgam reductions are: (a) complete reduction is
achieved in a few minutes; (b) the amalgam can be used repeatedly; and (c) no
blank correction is required as in the Jones redactor. The reduction potentials of
the saturated metal amalgams are as follows :

Zn^"*" -t-2e^Zn; —0.76 volt ■ ,

Cd2+-h2e^Cd; -0.40 volt ■ . •

Pb^+-|-2e^Pb; -0.13 volt

BiO + -f 2H + -b 3e Bi -t- HjO ; -f 0.32 volt

The most powerful reductant is therefore zinc amalgam, while bismuth amalgam
is the least reducing. The final reduction products obtained with these amalgams
for a few elements are collected in the table.

Liquid

amalgam

Iron

Titanium

Molybdenum Vanadium

Uranium

Tungsten

Zinc

Fe^ +

Ti3 +

V3-H

W2 4

Cadmium

Fe2 +

Ti3 +

V3 +

Lead

Fe^+

Ti3 +

V2 +

U++

W3 +

Bismuth

Fe^+

Ti3 +

Mo’+*

U4 +

* The exact product depends upon the pH of the solution,
t Some U^'*' is also formed.

The zinc amalgam is prepared by washing 15 g of pure, fine-mesh zinc shot (e.g.,
A.R.) with dilute sulphuric acid, and then heating for 1 hour on the water bath
with 300 g of mercury plus 5 cm^ of 1:4 sulphuric acid. (CAUTION. Mercury
vapour is highly poisonous; the operation must therefore be performed in a fume
cupboard with a good draught.) The whole is allowed to cool, the amalgam
washed several times with dilute sulphuric acid, and the liquid portion separated
from the solid by means of a separating-funnel. The solid is reserved for another
preparation of the amalgam. The liquid amalgam is preserved under dilute
sulphuric acid; reaction with the latter is very slow, and the same sample of
amalgam may be employed for several reductions. The other amalgams are

397

TITRIMETRIC ANALYSIS X, 146

tapping, and washed by decantation with dilute sulphuric acid. About 30 g of
silver in this form occupy a volume of 40-50 cm^ — sufficient to fill one reductor
tube. ‘

The necessary quantity of silver is introduced into the reductor above a small
plug of glass wool: by means of a glass rod flattened at one end, it is compressed
to as great an extent as necessary without restricting the free flow of solution
through the column. The reductor is rinsed with 100 cm^ of M-hydrochloric acid,
added in five equal, portions, each consecutive portion being allowed to pass
through the reductor to just above the level of the silver. . ■

The dark silver chloride coating which covers the silver of the upper part of the
reductor when hydrochloric acid solutions are employed moves farther down the
column in use, and when it extends to about three-quarters of the length of the
column, the reductor must be regenerated by the following method. The reductor
is rinsed with water and filled completely with 1 : 3-ammonia solution. The silver
chloride dissolves; after 10 minutes, the solution is rinsed out of the reductor tube
with water, followed by M-hydrochloric acid and is then ready for re-use. As a
precautionary measure, the ammoniacal solution of silver chloride should be
immediately acidified. The wastage of silver associated with this method of
regeneration may be avoided by filling the tube with sulphuric acid (O.IM) and
then inserting a rod of zinc with its lower end well buried in the silver; when the
reduction is complete (as evidenced by loss of the dark colour), the column is well
washed with water and is then ready for use.

Examples of the use of the silver reductor are given in Sections X, 110 and X,

111 .

X, 146. OTHER METHODS OF REDUCTION. Although as already
stated the use of metal amalgams, and in particular use of the Jones reductor or of
the related silver reductor, is the best method of reducing solutions in preparation
for titration with an oxidant, it may happen that for occasional use there is no
Jones reductor available, and a simpler procedure will commend itself. In
practical terms, the need is most likely to arise in connection with the
determination of iron, and the following reagents are amongst those most
commonly employed for the reduction of iron(III) to iron(II).

A. Tin(n) chloride solution. Many iron ores are brought into solution with
concentrated hydrochloric acid and the resulting solution may be readily reduced
with tin(II) chloride:

2Fe3+-t-Sn2+ =2Fe^+-t-Sn'^+ ■. . .

The hot solution (70-90 °C) from about 0.3 g of iron ore which should occupy a
volume of 25-30 cm^ and be 5-6M with respect to hydrochloric acid, is reduced
by adding concentrated tin(II) chloride solution dropwise from a separating-
funnel or a burette, with stirring, until the yellow colour of the solution has nearly
disappeared. The reduction is then completed by diluting the concentrated
solution of tin(II) chloride with 2 volumes of dilute hydrochloric acid, and adding
the dilute solution dropwise, with agitation after each addition, until the liquid
has a faint green colour, quite free from any tinge of yellow. The solution is then
rapidly cooled under the tap to about 20 °C, with protection from the air, and the
slight excess of tin(II) chloride present removed by adding 10 cm^ of a saturated
solution (ca. 5 per cent) of mercury(II) chloride rapidly in one portion and with
thorough mixing; a slight silky white precipitate of mercury(I) chloride should be
obtained. , ■

399

X, 146 QUANTITATIVE INORGANIC ANALYSIS

The small amount of mcrcur>(I) chloride in suspension has no appreciable
eUccl upon the oxidising agent used in the subsequent titration, but if a heavy
precipitate forms, or a grey or black precipitate is obtained, too much tin(li)
solution has been used; the results are inaccurate and the reduction must be
repeated. Finely divided mercury reduce.s permanganate or dichromate ions and
also slowly reduces Fe^ " ions in Ihe presence of chloride ion.

After the addition of the mcrcurylH) chloride solution, the whole is allowed to
stand for five minutes, then diluted to about 4tX) ciiv* and titrated with standard
potassium dichromatc solution (Section X, 103). or with standard permanganate
solution in the presence of ’preventive .solution’ (Section X, 93).

Blank runs- on the reagents should be carried through all the openiiions, and
corrections made, if necessary.

The concentrated .solution of tiulHl chloride i.s prepared by dissolving 12 gof
pure tin or 30 g of A.K. crystallised tiiuH) chloride (SnClj, 211^0) in KXIcin’’ of
concentrated hydrochloric acid and diluting to 200 crif* with water.

B. Reduction with sulphurous aciiL The solution must be feebly acid ( < 1 N)
and fairly dilute, say, 5U<) for 0.5 g of iron. If the concentration of the acid
exceeds 5,V, sulphurous acid wdl oxidise iron(U) solutions. Mydrochloric acid-
chloride soluiion.s are reduced more rapidly than sulphuric acid-sulphate
solutions. Kilher sulphur dioxide from a siplion of the liquid ga.s or fiesMy
prepared sulphurous acid solution or ammonium hydrogeiisulphite solution
may be used. The operation is best carried out in .i special all-glass wash bottleor,
if this Is not available, in a tlask liticd with a rubber stop[>er carrying two ‘wash-
bottle’ lubes.

Treat the hsdroehloric acid or sulphuric acid solution of the iron slowly and
with constant shaking witli dilute ammonia solution until a faint perntanent
precipitate Ls obtained. Dilute to about lOOcin'*, pass sulphur dio.xide through the
solution for 2-3 minutc.s, and then gradually heal to boiling, still continuing the
passage of tlie gas. When the solution is colourless (15-30 minutesl, replace the
sulphur dio.xide by a stream of w.isbcd carbon dioxide (from a Kipp’s apparatus
or cylinder), and boil vigorously until all the sulphur dioxide is expelled (20-30
minutes) as shown by p.i.ssing tlte escaping gas for 30 seconds through dilute
sulphuric acid containing 2 drops of O-lA’-permanganate, AIKnv the solution to
cool in a stream of carbon dioxide, add more acid, and titrate with a .standard
solution of a suitable oxidising agent.

A simpler method is to place the acidified iron .solution in a conical liask, add
dilute ammonia solution slowly until a faint permanent precipitate is obtained,
and then add either 25 cm^ of a freshly prep.ircd saturated solution of sulphur
dioxide or e.xcess of frestily prepared ammonium hydrogensulphite solution; in
the latter case a little dilute sulphuric acid is added. A small funnel is placed in the
mouth of the Ilask, and the mi.\ture boiled for 30 minutes. All the sulphur dio.xide
will then have been expelled. Cool the .solution in an atmosphere of carbon
dioxide, add 10 cur' of dilute sulphuric acid (1:6), and titrate at once with a
standard solution of the oxidising agent.

It has lK*cn found that a higher acidity can be tolerated (<2<V in H.SOJ.and
the reaction is accelerated, in the presence of thiocyanate ion. The procedure for
the thioeyanate-acceleralcd reduction is ;is follows. Add 10 cm^ of O.LV-
potassium thiocyanate to the .solution of iron(U 1), w-hich should be less than M in
sidphuric acid. Saturate the solution in the cold with sulphur dio.xide, or add 50
cm^ of Ireshly prepared sulphurous acid solution. Heat slowly to the boiling

400

TITRIMETRIC ANALYSIS X, 147

point. The solution rapidly becomes colourless or pale yellow. Displace the
sulphur dioxide with carbon dioxide or nitrogen, cool, and add 10 cm^ of O.IM-
mercury(II) nitrate solution (to complex the thiocyanate). Titrate with standard
permanganate (or dichromate) solution in the usual way.

Members of the hydrogen sulphide group of metals must be absent. If present,
they must be removed first. Titanium and chromium are unaffected by the
treatment; vanadium(V) is reduced to vanadium(IV).

G Reduction wth hydrogen sulphide. The method is not frequently
employed. A typical procedure is as follows. The solution for reduction (ca. 200
cm^) should be about 0.5M in sulphuric acid. Heat to boiling, and pass a stream
of washed hydrogen sulphide until the solution is saturated. Remove. from the
source of heat, and continue to pass the gas for a further 15 minutes. Boil the
solution down to about 50 cm^ during 30-60 minutes while a stream of oxygen-
free carbon dioxide is passed through. The solution is allowed to cool in a stream
of the gas, diluted to 200 cm^ with distilled water, and titrated with standard
permanganate solution. The precipitated sulphur is coagulated during the
concentration, and usually need not be removed before titration.

Hydrogen sulphide is sometimes used for the reduction of Fe(III) to Fe(n)
because of its selectivity. Copper is precipitated as sulphide and is filtered off;
vanadium(V) is reduced to vanadium(IV), which does not interfere in the
subsequent titration provided dichromate is used. Molybdenum is largely
precipitated and is filtered off; the hydrogen sulphide is boiled out of the filtrate, a
few drops of permanganate solution are added to re-oxidise the reduced
molybdenum, hydrogen sulphide passed again, and the remaining molybdenum
sulphide separated by filtration.

The reagents listed above can be applied to the reduction of many other ions in
addition to Fe^'*', and there are also a number of other substances which can be
employed as reducing agents; thus for example hydroxylammonium salts are
frequently added to solutions to ensure that reagents do not undergo atmospheric
oxidation, and as an example of an unusual reducing agent, phosphorous acid
may be used to reduce mercury(II) to mercury(I); see Section X, 134.

X, 147. References

1. International Union of Pure and Applied Chemistry (Aug. 1974). Information
Bulletin. No. 36.

2. M. L. McGlashan (1971). Physico-Chemical Quantities and Units. 2nd edn., p. 45.
London; Royal Institute of Chemistry.

3’. A. I. Vogel (1958). Elementary Practical Organic Chemistry. Pt. III. Quantitative
Organic Analysis. London; Longmans Green and Co.

4. FLA. Cotton and G. Wilkinson (1972). Advanced Inorganic Chemistry. 3rd. edn.
London; Interscience Publishers.

5. S. F. A. Kettle (1969). Co-ordination Compounds. London; T. Nelson and Sons Ltd.

6. G. Schwarzenbach and H. Flaschka (1969). Complexometric Titrations. 2nd edn.
London; Methuen and Co.

7. R.Pribil and V. Vesely (1961). Chemist-Analyst, 50, 100.

8. T. S. West. Complexometry (1969). 3rd edn. Poole; BDH Chemicals Ltd.

9. Society of Dyers and Colourists (1956). Colour Index. 2nd edn. Bradford

10. R. A. Close and T. S. West (1960). Talanta, 5, 221.

■ ?■ Woodward and H. N. Redman (1973). High-precision Titrimetry. London; Society
for Analytical Chemistry.

401

X, 148 QUANTITATIVE INORGANIC ANALYSIS

12. IT N. Wilson (195 1 ). Analysl, 76. 65.

1 3. C. C. Reilly aiui W. W. Portcrlicld (1956). Anulyikal Clwiubiry. 28, 443.

14. L. W. Amlrcws (1903). J. Ant. Ckvnt. Sac., 25, 756.

1 5. G. S. J.imicson ( 1 926). Valumdric ladaw .Method!. New York; Reinbold.

16. C. M. Ellis and A, I. Voycl ( 19S<>). .-UudyA, 81, 693.

17. H. W. Smilh and ,\1. L. Parsons (1 973). J. Clicm. Ed.. 50, 679.

18. A. R. .Morrison (1972). Lab. I'racticc, 21 , 726,

19. Inlern.ilional Union of Pure and Applied Chemistry (Sept. 1975). Informauan
liulteiin. No. 45.

20. -A. Rinehoin (1963). Comjdcxationm .'inalyikal Chemistry. New York: Intersdence,

21. Slahiiiiy Conttaius of .Metul-hm CompUwes. Chemical Society Special Publications
Nos. 17 and 25. London.

22. Lahoratury ll’aitc Duptnal 3(«.*)«o/t,1969). 2ndcd. W'ashinpton D.C. ; Manufactur-
ing Chemists A-ssocialion. (rev. edit. 1974).

23. P. J. G.iston (196-}). The Care. Uandwni and Disposal a/ Dangerous Cheinicab.
Aberdeen; Northern Publi'-hcfs Ltd.

X, 148. Sclcc(ccl bibliography

1. D. Bciteridge and 11. E. llailain (1972). .Modern .inalyticat .Methods. London; The
Chemical Society.

2. E. Bishop (1972). Indicators. Oxford; Pcrganion Press Ltd.

3. N. U. Eurnun (1962). Standard .Methods of Chemical .-inahsis. 6ih cdn. Princeton,
N.J.; Van Nostrand.

4. G. Jandcr (1956). .S'euere mass.inalitiuhen .Methoden. Stuttg.irt; rcrdiiund Enkc
Vcrl.ii;.

5. I. M. Koltholf and V. .A Slcngcr, IVliimetric .-Irmlviis. Vol. I (1942). Vol. 11 (1947),
1. .M. Koltholf and R Belcher. Vol. 111(1957). New York; Inicrscicncc Publishers.

6 1. .M Koltholf and P. J Living ( 1961) Treatise on .Inal} lical Chemistry. KevrYotk;
Interscicnee Publisher .s.

7. L. Mciles(1963). Handbook of .Inahtteal Chenmtry. New York: ,\lcGrasv-lliII,

8. R. Pribil (1972). Analytical .Ippite.itions of EDT.i and Related Compounds. Oxford:
Pergaroon Press Lid.

9. G. Schvvar/cnbach and 11. Masch.ka (1969). Comple.xomelric Titrations. 2nd cdn.
London; Methuen and Co.

10. W. Wagner and C. J. Hull (1971). Inorgante Tiinmetric .tnalysh. Nesv York; Marcel
Dekkerinc.

11. C. L. Wilson and O. W, Wilson (1962). Camprehen.\ite Analylicid Chemistry.
Amsicrdain; Elsevier.

12. L. F. Hainiiton. S. G. Simpson and D. W. Ellis (19ti9). Calculations of Analytical
Chemistry. 7tli cdn. New York ; .McGraw-Hill,

Many ol the general textbooks listed in Chapter I wdl also be relevant.

402

CHAPTER XI GRAVIMETRY

XI, 1. INTRODUCTION TO GRAVIMETRIC ANALYSIS. Gravimetric \
analysis or quantitative analysis by weight is the process of isolating and \
weighing an element or a definite compound of the elerrient in as pure a' form as
possible. The element or compound is separated from a weighed portion of the
substance being examined. A large proportion of' the determinations in
gravimetric analysis is concerned with the transformation of the element or
radical to be determined into a pure stable compound which can be readily r-
converted into a form suitable for weighing. The weight of the element or radical
may then be readily calculated from a knowledge of the formula of the (
compound and the atomic weights of the constituent elements.

The separation of the element or of the compound containing it may be
effected in a number of ways, the most important of which sire : (a) precipitation
methods, (d) volatilisation or evolution methods, (c) electroanalytical methods,
and (d) extraction and chromatographic methods. Only (a) and (b) will be
discussed in this chapter : (c) is considered in Part E, and (d) in Part C.

It may be mentioned at this stage that the great advantage of gravimetric over
titrimetric analysis is that the constituent is isolated and may be examined for the
presence of impurities and a correction applied, if necessary ; the disadvantage of
gravimetric methods is that they are generally more time-consuming.

XI, 2. PRECIPITATION METHODS. These are perhaps the most import-
ant with which we are concerned in gravimetric analysis. The constituent being
determined is preci pitate d from solution in a form whi ch is so slightly soluble t haL
no appreciable loss occufT" whenTheTJrecipitate'i^separated by ffltration and
weighed. Thus in the determination of silver, a solution of the substance is treated
with an excess of sodium or potassium chloride solution, the precipitate is filtered
off, well washed to remove soluble salts, dried at 130-1 50 °C, and weighed as
silver chloride. Frequently the constituent being estimated is weighed in a form
other than that in which it was precipitated. Thus magnesium is precipitated, as
ammonium magnesium phosphate Mg(NH4)P04,6H20, but is weighed, after
ignition, as the pyrophosphate Mg2P207. The factors which determine a
successful analysis by precipitation are;

1. The precipitate must be so insoluble that no appreciable loss occurs when it is
collected by filtration. In practice this usually means that the quantity
remaining in solution does not exceed the minimum detectable by the ordinary
analytical balance, viz., 0.1 mg.

403

XI, 3 QUANTITATIVE INORGANIC ANALYSIS

2. The physiiMl nature of the precipitate must be such that it can be readily
se[ 3 arated from the solution by filtration, and can be washed free of soluble
impurities. Tlicse conditions require that the particles are of such size that they
do not pass through the filtering medium, and that the particle size is
unalfected (or, at least, not diminished) by the washing process.

3. The precipitate must be consertible into a pure substance of definite chemical
composition; this may be elfccted either by ignition or by a simple chemical
opcnition. such as evaporation, with a suitable liquid.

Factor 1 . which is concerned with the completeness of precipitation, has
already been dealt with in connection with the solubility product principle
(Section.s H, 8 and 9), and the inlluence upon the solubility of the precipitate of(i)
a salt with a common ion, (ii) salts with no common ion. (iii) acids and bases, and
(iv) temperature.

It was assumed ihrougliout that the compound which separated out from the
solution was chemically pure, but tills is not always the case. The purity of the
precipitate depends inWr ului upon the substances present in solution both before
and after the addition of the reagent, and also upon lhee.\act e,\perimental con-
ditions of precipitation. In order to understand the intluenee of these and other
factors, it will be necessary to give a short account of ilic propcrtie.s of colloids.

Problems svhich arise with certain precipitates include the coagulation or
llocculation of a colloidal dispersion of a finely divided .solid to permit its
filtration and to prevent us repepiisation upon washing the precipitate. It is
therefore desirable to understand tiic basic principles of the colloid chemistry of
precipitates.

XI, 3. THK COLLOIDAL ST.VrE. The colloidal stale of matter is distin-
guished by a certain range of particle size, as a consequence of which certain
characteristic properties become apparent. Hefore discussing these, mention
must be made of the various units whicii arc employed in expressing small
dimensions. The most imponani of these are:

1pm --- 10" •’mm I nm 10"“ nun

I Angstrom mm A -= 10' metre 10" ’ mm - O.l nni
Colloidal properties are, m general, e.xhibiied by substances of particle size
ranging between 0. 1 pm and I nm. Ordinary qiKiiuitalivc filter paper will reltiin
particles up to a diameter of about 10' * mni or^lU/mi, so that colloidal solutions
in this respect behave like true soluiions (size of molecules is of the order of
0. 1 nm or 1 0' •‘’cml The limit of vision under tiie microscope is about 0.2 pm. Ifa
powerfiiTBeam ol light is passed through a colloidal solution and the .solution
viewed at right angles to the incident light, .i scattering of light is observed. This is
the so-called Tyndall elfect. Truesolution.s. i.e., ihu.se witirparliclcs of molecular
dimensions, do not c.xhibit a Tyndall cOect, and .irc said to be ‘optically empty'.
Use is made ol the I yiidall ellect in tlic uhra-microscope; here the Tyndall cone
or beam is observed in a microscope whicli is situated at right angles to the path
of the incident light. The dilTraction images arc ilurs seen, and it is possible to
observe the light scallercd by each parliele separately. The limit of visibility
under the ultra-microscope is about 10 nm.

Uy the use of X-rays the physical structure of the smallest unit of colloidal
substances may be ascertained, it has been found that most substances consist of
minute crystalline particles; ;i few, such as silica and lin(IV) oxide, arc
amorphous. An intermediate .stage is also possible; a gradual development of

•104

GRAVIMETRY XI, 3

crystalline particles may occur with some amorphous substances upon ‘ageing’
or with suitable treatment, such as digestion with hot water or solutions of
electrolytes (Section XI, 5). ■ \

An important consequence of the smallness of the size of the particles is that the
ratio of surface area to weight is extremely large. Phenomena, such as adsorption,
which depend upon the size of the surface will therefore play an important part
with substances in the colloidal state. Table XI, 1 clearly shows the influence of
particle size in connection with a 1 -cm cube decimally divided.

Table XI, 1 Increase in number and total surface of particles
as a one centimetre cube is decimally divided

Number of particles

Length of edge in cm

Total surface in cm’

10®

10'^

6x 10’

10‘^

10"* (= 1/jm)

6x10*

10‘®

10"®

6x 10®

10^‘

10"’ (= 1 nm)

6x 10’

10^“

10"® (= 1 A)

6x10®

The characteristic properties of most types of colloidal particles encountered
in inorganic analysis are :

(a) they exhibit a Tyndall effect when viewed with proper illumination (see,
however, the table below) ;

{b) they may be separated from true solutions of substances by means of a
collodion or parchment membrane, i.e., by the process of dialysis;

(c) they possess electrical charges since they migrate under the influence of a
suitable potential gradient;
id) they possess a very large surface area.

For convenience, we may divide colloids into two main groups, designated as
lyophobic and lyophilic colloids. The chief properties of each class are summarised
in the following table, although it must be emphasised that the distinction is not
an absolute one, since some gelatinous precipitates (e.g., aluminium and other
metallic hydroxides) have properties intermediate between those of lyophobic
and lyophilic colloids.

Lyophobic colloids

Lyophilic colloids

I ■ The dispersion (or sols) are only slightly
viscous. Examples: sols of metals, silver
halides, metallic sulphides, etc.

2. A comparatively minute concentration of an
electrolyte results in flocculation. The change
is, in general, irreversible; water has no effect
upon the flocculated solid.

3. Lyophobic colloids, ordinarily, have an
electric charge of definite sign, which can be
changed only by special methods.

4. The ultra-microscope reveals bright particles
in vigorous motion (Brownian movement).

1 . The dispersions are very viscous ; they set to
jelly-like masses known as gels. Examples :
sols of silicic acid, tin(IV) oxide, gelatin.

2. Comparatively large concentrations of
electrolytes are required to cause
precipitation (‘salting out’).The change is, in
general, reversible, and reversal is effected by
the addition of a solvent (water).

3. Most lyophilic colloids change their charge
readily, e.g., they are positively charged in
acid medium and negatively charged in an
alkaline medium.

4. Only a diffuse light cone is exhibited under
the ultra-microscope.

The process of dispersing a gel or a flocculated solid to form a sol is called
peptisation.

The stability of lyophobic colloids is intimately associated with the electrical

405

XI, 3 QUANTITATIVE INORGANIC ANALYSIS

charge on the particles.'' Thus in the formation of an ar.scnic(in) sulpiiide so) by
precipitation with hydrogen sulphide in acid solution, sulphide ions are primarily
adsorbed (.since every precipitate has a tendency to adsorb its own ions), and

some hydrogen ions are secondarily adsorbed.
Tile hydrogen ions or other ions which are
secondarily adsorbed have been termed coun-
ter ions. Thus the so-called electrical double
layer is set up between the particles and the
solution. An arsenic) HI) .sulphide particle is
represented diagrammatically in Fig XI, I.
The colloidal particle of arsenic(lll) sulphide
has a negatively charged surface, with po-
sitively charged counter ions which impart a
positive charge to the liquid immediately
surrounding it. If an electric current is passed
through the .solution, the negative particles will move towards the anode: the
speed is comparable with that of electrolytic ions.Thc electrical conductivity of a
sol is, however, quite low because the number of current-earry’ing particles is
small compared with that in a solution of an electrolyte at an appreciable
concentration ; the large cliargc carried by the colloidal particles is not sufficient
to compensate for their smaller number.

If the electrieal double layer is devlroycd. the sol is no longer stable, and the
particles will flocculate, thereby reducing the large surface area. Thus if barium
chloride solution is added, barium ioas are preferentially adsorbed by the
particles; the charge distribution on the surface is disturbed and the particles
flocculate. After llocculaiion. it is found that the dis|Krsion medium is acid owing
to the liberation of the hydrogen counter ioas. It appears that ions of opposite
ciiarge to those primarily arborbed on the surface arc necessary for cogulalion.
The minimum amount of electrolyte necessary to cause llocculaiion of thecolloid
is called the flocculation or coagulation value. It has been found that the latter
depends primarily upon the valency of the ions of the opposite charge to that on
the colloidal particles: the nature of the ions has some influence also. This is
clearly shown by the results colieeted m Table .\I. 2.

1 able XI, 2 Coagululiun values iii inilli-muls of cuagulatiiig ioii per litre

Ncealivc a

S.ilt

rscnic(Ill) sulpbi Je sol.

Coag, value

I’usiti V e li) dratci) irun(l I i ) u side sub

ii.iU Coag. Value

AICI,

0,062

KjFciCN'L

O.Oo

AI,(SO.)j

0.074

K.FctC.M.,

OIW

Fed,

0.1.16

K,SO,

CaCf,

0.649

K.Cr.O.

0.19

HaCT

0.091

K.Oth

0.13

MgCL

0717

KjC-O.

0 24

Ua(NO,),

0,6S7

KDtO,

.11

KCl

49.5

K.S('.N

NaCI

.51,0

KCl

101

UCI

5S.4

KNOj

111

KNO,

50.U

KUr

1,1S

IlG

30.S

1S4

• Ljophilic eolloidi arc nuinly slabiliscJ b> solvation.

HMt'

K'U*

Fig. .XI, I

-106

GRAVIMETRY XI, 4

If two sols of opposite sign are mixed, mutual coagulation usually occurs
owing to the neutralisation of charges. The above remarks apply largely to.
lyophobic colloids. Lyophilic colloids are generally much more difficult to
coagulate than lyophobic colloids. If alyophilic colloid, e.g., of gelatin, is added
to a lyophobic colloid, e.g., of gold, then the lyophobic. colloid appears to be
strongly. protected against the flocculating action, of electrolytes. It is probable
that the particles of the lyophilic colloid are adsorbed by the lyophobic colloid
and impart their own properties to the latter. The lyophilic colloid is known as a
protective colloid. This explains the relative stability produced by the addition of
a little gelatin to the otherwise unstable gold sols. For this reason also, organic
matter, which might form a protective colloid, is generally destroyed before
proceeding with an inorganic analysis.

During the flocculation of a colloid by an electrolyte, the ions of opposite sign
to that of the colloid are adsorbed to a varying degree on the surface ; the higher
the valency of the ion, the more strongly is it adsorbed. In all cases, the precipitate
will be contaminated by surface adsorption. Upon washing the precipitate with
water, part of the adsorbed electrolyte is removed, and a new difficulty may arise.
The electrolyte concentration in the supernatant liquid may fall below the
coagulation value, and the precipitate may pass into colloidal solution again.
This phenomenon, which is known as peptisation, is of great importance in
quantitative analysis. By way of illustration, let us consider the precipitation of
silver by excess of chloride ions in acid solution and the subsequent washing of
the coagulated silver chloride with water; the adsorbed hydrogen ions will be
removed by the washing process and a portion of the precipitate may pass
through the filter. If, however, washing is carried out with dilute nitric acid, no
peptisation occurs. For this reason, precipitates are always washed with a
suitable solution of an electrolyte which does not interfere with the subsequent
steps in the determination.

The adsorptive properties of colloids find anumber of applications in analysis,
e.g., in the removal of phosphates by hydrated tin(IV) oxide in the presence of
nitric acid, in the use of adsorption indicators Section X, 30, C, in the qualitative
detection and colorimetric determination of elements and radicals with many
organic reagents (for example, magnesium with Titan yellow. Section XVin, 22 ,
A).

XI, 4. SUPERSATURATION AND PRECIPITATE FORMATION. The
solubility of a substance at any given temperature in a given solvent is the amoimt
of the substance dissolved by a known weight of that solvent when the substance
IS m equilibrium with the solvent. The solubility depends upon the particle size,
when these are smaller than about 0.01 mm in diameter ; the solubility in creases '
grratly the smaller the particles, owing to the increasing role played by surface
effects (compare'Table'XI, 1). (The definition of solubility given above refers to
particles larger than 0.01 mm.) A supersaturated snlntinn is one that contains a
greater concentration of solute than cqrresponds to the equilibrium solubility at
the temperature under consideration. Supersaturation is therefore an unstable
state which may be brought to a state of stable equilibrium by the addition of a
crystal of the solute (‘seeding’ the solution) or of some other substance, or by
mechanical means such as shaking or stirring. The difficulty of precipitation of
ammonium magnesium phosphate will at once come to mind as an example of
supersaturation.

According to von Weimarn supersaturation plays an important part in

407

Xr, 4 QUANTITATIVE INORGANIC ANALYSIS

determining the partiele -size ofa precipitate. He deduced that the initial velocity
of precipitation is proportional to (Q--S)IS, where Q is tlie total concentration
oflhe .sulwlance that is to precipitate, and S is the equilibrium .solubility ; (g-J;
will denote the supersatunition at the moment precipitation commences. The
e.xpression applies appro, vimaiely only when Q is large ;is compared with S. The
influence of the degree of.super.saluration is well illustrated by von Weiinam's
re.suU.s for the fortnation of barium sulphate from solutions of barium
thiocyanate and numgaiiese sulphate respectively, fhese tire collected in Table
XL 3. The results clearly show that the particle size ofa precipitate decreases with

Table .XF, 3 .Separation of ilaSO,; at sarious degree-s of supersaturation
(son Weimani)

ContcnlraliuH

oftesgciib

r)pc of

7.V

17JU00

A piccipil.itcis foir.icO, jiul piat'Uo.illj ll;c a hole of

(iic'.i.a.ru ii:iajv>bili'cJ; tbc aHiiainiiii; vt-iwl wn he
iMibtjiii ibctyiiicnu running uul. I lie gel u un>t.rl!!c. ar.4 gro Aiii
of the large eo-tah ai ihc capen-.c of ‘-mall oim A \er> rapid;
alter a lew luiuts ilie piccipilate hceoir,t» iipaque.

.VA’

75000

Gclauiiciua liliiu formed; Isxoiue lutbiJ alter one ininuic.

25tKK|

I’uiii.ira pieapiiate u cuuly .md of eolUnJal diisicniwtis.
i’atlic!e>appc.ir.i.a jMinls .il a inagniliealionof liWx.

o.os.y

I’tun.ityptreapiiatcoonH.iaof fealbeiy and vlar-htapcjcryslal
siicleloiu

o.twi.v

155

i'lCeipilalrcon'.iiis ofeuiupael crjitai •.kclcumi.

Cl. 0.00 i.v

S.shiUoii h^comet opalcwciil duimg liiM 5 tsitnuicip. and
ptceipiUtiotieamtim.cs lor 2 -5 lu.'ura. .Alter that uineeiysub
lia'.c a lenatli iil'O W5 m.'il

t‘reeipitaleapi>e.i[a alter about a ir.onlh. .At tbeendof ai\
mcnlhi. lhelc!i.eih .'f thclai,;nl etsst.il.a ii aha.‘Ul O.Ud mm and
ibeir htcadlh 0 O! 5 mm

increa.sing concentration of the reactants, h'or the production of a crystalline
precipitate, fur which the adsorption errors will be least and filtration will be
easiest, {Q-S);S .should be as snutll as pu.ssible. There is obviously a practical
limit to reducing (0 - .ST.'.S by making Q very small, since for a precipitation to be
ol value in analysis, it must be complete in a comparatively .short time and the
volume.s of solutions involved must not he tot) large. There is. however, another
method which may be used. vi/.. that of incre.ising ,V. For e.xample, barium
.sulphate is about fifty times more soluble in 2,l/-hydroeliloricacid than in water:
it 0.053/ solutions of barium chloride and sulphuric acid are pa’pared in 2.1/
boiling hydrochloric acid and the .solutions mi.xed. a typical co’Stalline pre-
cipitate of barium .suliihatc is slowly formed (Refs. 1 and 2).

.Applications of the above conceptions are to be found in the following
recognised procedures in gnivimetric analysis:

1. Precipitation is usually c;irried out in hot solutions, since the solubility
generally increases with rise m temper.tlure.

2. Precipitation is cllected in dilute solution and the reagent is added slowly and
with thorough stirring. The slow addition results in the first particles
precipitated acting as nuclei which grow a.s further material precipitates.

3. A suitable reagent is often added to increase the solubility oflhe precipitate
and thus lead to larger primary particles.

408

GRAVIMETRY XI, 5,

4. A procedure which is commonly employed. to prevent supersaturation from
occurring is that of precipitation from homogeneous solution. This is achieved
by forming the precipitating agent within the solution by means of an
homogeneous reaction at a similar rate to that required for precipitation of the
. species.

XI, 5 THE PURITY OF THE PRECIPITATE. CO-PRECIPITATION.
When a precipitate separates from a solution, it is not always perfectly pure: it
may contain varying amounts of impurities dependent upon the nature of the
precipitate and the conditions of precipitation. The contamination of the
precipitate by substances which are normally soluble in the mother liquor is
termed co-precipitation. We must distinguish between two important types of co-
precipitation. The first is concerned with adsorption at the surface of the particles
exposed to the solution, and the second relates to the occlusion of foreign
substances during the process of crystal growth from the primary particles.

With regard to surface adsorption, this will, in general, be greatest for
gelatinous precipitates and’ least for those of pronounced macro-crystalline
character. Precipitates with ionic lattices appear to conform to the
Paneth-Fajans-Hahn adsorption rule, which states that the ion that is most
strongly adsorbed by an ionic substance (crystal lattice) is that ion which forms
the least soluble salt. Thus on sparingly soluble sulphates calcium ions are
adsorbed preferentially over magnesium ions because calcium sulphate is less
soluble than magnesium sulphate. Also silver iodide adsorbs silver acetate much
more strongly than silver nitrate under comparable conditions, since the former
is the less soluble. The deformability of the adsorbed ions and the electrolytic
dissociation of the adsorbed compound also have a considerable influence; the
smaller the dissociation of the compound, the greater is the adsorption. Thus
hydrogen sulphide, a weak electrolyte, is strongly adsorbed by metallic sulphides.

The second type of co-precipitation may be visualised as occurring during the
building up of the precipitate from the primary particles. The latter will be subject
to a certain amount of surface adsorption, and during their coalescence the
impurities will either be partially eliminated if large single crystals are formed and
the process takes place slowly, or, if coalescence is rapid, large crystals composed
of loosely bound small crysals may be produced and some of the impurities may
be entrained within the walls of the large crystals. If the impurity is isomorphous
or forms a solid solution with the precipitate, the amount of co-precipitation may
be very large, since there will be no tendency for elimination during the ‘ageing’
process. The latter actually occurs during the precipitation of barium sulphate in
the presence of alkali nitrates ; in this particular case X-ray studies have shown
that the abnormally large co-precipitation (which may be as high as 3.5 per cent if
precipitation occurs in the presence of high concentrations of nitrate) is due to
the formation of solid solutions. Fortunately, however, such cases are compara-
tively rare in analysis.

Appreciable errors may also be introduced by post-precipitation. This is the
precipitation which occurs on the surface of the first precipitate after its
tormation. It occurs with sparingly soluble substances which form super-
saturated solutions; they usually have an ion in common with the primary
precipitate. Thus in the precipitation of calcium as oxalate in the presence of
magnesium, magnesium oxalate separates out gradually upon the calcium
oxalate; the longer the precipitate is allowed to stand in contact with the solution.

409

-.XlOP QUANTITATl VF. INORGANIC ANALYSIS

. ■

the greater is the error due to this cause. A similar eflccl is observed in the
precipitation of copper or mercuryCII) sulphide in 0..1 A/-hydroehloric acid in the
presence of .tine ion.s; /inc sulphide is slowly posl-precipiUited.

Post-precipitation ditfers from co-precipitation in several respects:

{a) The comainination inerca.ses with the lime that the precipitate is left in
contact with the moiiier liquor in ptwi-precipiiation, but usually decreases in
co-prccipilation.

(h) With post-precipitation, contamination increases the faster the solution is
agitated by eiilier mechanical or thermal itteans. The reverse is usually true
with co-precipitation,

(e) The magnitude of contamination by post-prccipitation may be much greater
than in co-prccipitalion.

It is convenient to consider now the inllucnce of digestion. This is usually
carried out by allowing tiie precipitate to stand for 12-2-t hours at room
temperature, or sometimes by warming the precipitate for some time, in contact
with the liquid from which it was formed; the object is. oj cmirs.c,„bi obtain
cqiupkie precipitation in a form which can he readily fdiered. During the process
of digestion or of the ageing of precipitates, at least two changes oceur. The \ery
small particles, which have a gre.ilcr solubility titan the larger ones, will, after
precipitation has occurred, tend to pa.ss into solution, and will ultimately rc-
deposit upon the larger particles; co-precipitation on the minute particles is thus
eliminated and tlie total co-precipitation on the ultimate precipitate reduced. The
rapidly formed cry stabs arc probably of irregular shape and possess a compara-
tively large surface: upon digestion tiiesc tend to become more regular in
character and also more dense, thus resulting in a decrease in the area of the
surface and a consequent rcductum of adsorption. The net result of digestion is
usually to reduce the extent of eo-precipitation and to mcrea.se the si« of the
paiiicic.s, rendering filtration e.isier.

XI, 6. CONDI TIONS (JF TKIXTPTTA TION, No universal rules can be
given which are applicable to ail eases of precipitation, but. with the aid of an
inteliigenl application of the f.icis enumerated in the foregoing paragraphs, a
number of fairly general rules may be stated ;

. ^ 1. Precipitation .should be Carried out in dilutesoluiion, due regard being paid

to the .solubility of the precipitate, the time required for filtration, and the
' subsequent operations to be e-arried out with llie liltrate. This will minimise the
errors due loco-precipitation.

^ , 2. 1 he reagents should be mi.xed slowly and with constant stirring. Thi,s will

keep the degree of supcrsaturatioii small and vvill a.s,sisi the growth of largc^
crystals. A slijglu c.vcess of the reagent is all that is generally required; in
exceptional ca,scs a large c.xccss may be necessary. In some instances the order of
mi.xing the reagents may be important. Frceipiiation may be ciTected under
conditions wliich increa.se the soltibiltty of the precipitate, tiiU-S further reducing
the degree of supcrsaturatioii (compare Section XI, 5J.

3. Precipitation is cITcctcd in hot solutions, provided the solubility and the
stability of the precipitate permit. Either one or both of the solutions should be
lieated to just below tlie boiling point or other more favourable temperature. At
the higher temperature: (o) the solubility is increased with a consequent
reduction in the degree ol supersaturalion, (h) coagulation is assisted and sol

410

GRAVIMETRY XI, 7

formation decreased, and (c) the velocity of crystallisation is increased, thus
leading to better-formed crystals.

4. Crystalline precipitates should be digested for as long as practical,
preferably overnight, except in those cases where post-precipitation may occur.
As a rule, digestion on the steam bath is desirable. This process decreases the
effect of co-precipitat ion and gives mo re readily filterable.precipitates. Digestion
has little effecTupoifamorphous or gelatinous precipitates.

5. The precipitate should be washed with the appropriate dilute solution of an
electrolyte. Pure water may tend to cause peptisation. (For theory of washing, see
Section XI, 8 below.)

6. If the . precipitate is still appreciably contaminated as a result of
co-precipitation or other causes, the error may often be reduced by dissolving
it in a suitable solvent and then reprecipitating it. The amount of foreign
substance present in the second precipitation will be small, and consequently the
amount of the entrainment by the precipitate will also be small.

XI, 7. PRECIPITATION FROM HOMOGENEOUS SOLUTION. The
major objective of a precipitation reaction is the separation of a pure solid phase
in a compact and dense form which can be filtered easily. The importance of a
small degree of supersaturation has long been appreciated, and it is for this
reason that a dilute solution of a precipitating agent is added slowly and with
stirring. In the technique known as precipitation from homogeneous solution the
precipitant is not added as such, but is slowly generated by a homogeneous
chemical reaction within the solution. The precipitate is thus formed under
conditions which eliminate the undesirable concentration effects which are
inevitably associated with the conventional precipitation process. The pre-
cipitate is dense and readily filterable; co-precipitation is reduced to a minimum.
Moreover, by varying the rate of the chemical reaction producing the precipitant
inhomogeneous solution, it is possible to alter further the physical appearance of
the precipitate— the slower the reaction, the larger (in general) are the crystals
formed.

Many different anions can be generated at a slow rate ; the nature of the anion
IS important in the formation of compact precipitates. It is convenient to deal
with the subject imder separate headings.

(a) Hydroxides and basic salts. The necessity for careful control of the pH
has long been recognised. This is accomplished by making use of the hydrolysis
of urea, which decomposes into ammonia and carbon dioxide as follows ;

C0(NH2)2+H20 = 2NH3+CO2

Urea possesses negligible basic properties (A* =1.5x10“ is soluble in water
and its hydrolysis rate can be easily controlled. It hydrolyses rapidly at 90-100 °C,
and hydrolysis can be quickly terminated at a desired pH by cooling the reactioii
nuxture to room temperature. The use of a hydrolytic reagent alone does not
result in the formation of a compact precipitate; the physical character of the
precipitate will be very much affected by the presence of certain anions. Thus in
the precipitation of aluminium by the urea process, a dense, precipitate is
obtained in the presence of succinate, sulphate, formate, oxalate, and benzoate,
but not in the presence of chloride, chlorate, perchlorate, nitrate, sulphate,
chromate, and acetate. The preferred anion for the precipitation of aluminium is
succinate. It would appear that the main function of the ‘suitable anion’ is the

411

XI. 7 quantitative inorganic analysis

formation of a basic salt which seems responsible for the production of a
compact precipitate. The pH of the initial .solution must be appropriately
adju.sied.

The following are suitable anions for urea precipitations of some metals;
sulphate for gairium. tin, and titanium ; formate for iron, thorium, and bismuth;
succinate for aluminium and zirconium.

The urea method generally resuits in the deposition on the surface of the
beaker of a thin, tenacious, and somewhat transparent film of the basic .salt. This
film cannot be removed by scraping with a ‘policeman'. It is dissolved by addinga
few cm-* of hydrochloric acid, covering the beaker with a clock glass, and
retluxing for 5-10 minutes; tiie small amount of metallic ion is precipitated by
ammonia solution and filtcr.s readily through the .same lilter containing the
previously precipitated basic salt.

The urea hydroly.sis method may be applied also to:

(i) the precipitation of barium as barium chromate in the presence of
ammonium acetate;

(ii) the precipitation of large amounts of nickel as the dimcthylglyo.ximate:and

(iii) the precipitation of aluminium as the oxinate.

(/)) I’lmpluiie.'i. Insoluble phosphates m;iy be precipitated with phos-
phate ion derived from irimethyi or triethyl phosphate by stepwise hydrolysis.
Thus l.S.U-.sulphiiric acid containing zirconyi ions and trimeihyl phosphate on
heating gives a dense precipitate of variable composition, which is ignited to and
weighed as the pyrophosphate ZrPjO-.

Mciaphosphonc acid may also be used; it hydrolyses in wann acid solution
forming orlhoplu)s|)horic acid. Thus bismuth may be precipitated as bismuth
pho.sphate in a dense, crystalline form.

U') Oxuliitcs. Urea may be employed to raise the pi! of an acid solution
containing hydrogenoxaiate ion IIC-O^' . ihu.s alfording a method for the slow
generation of oxidate ion. Calcium oxalate may thus be prccipit.ited in a dense
form;

C0(NH,), + 2nc:;0^ rll,0 „ 2NIU' f CO, r2Cj.Or"

Dimethyl and diethyl oxalate can be hydrolysed to serve as reagents for oxalate
ion;

{C\H,)iCi0.j-)-2H,0 - 2CT!jOH.f2H‘ vC;0,- •

Diethyl oxalate i.s usually prefer'cd because of its slower rate of hydrolysis.
Satisfactory results are obtaine<’ m the precipitation of calcium, magnesium, and
zinc: thorium is precipitated i.iiiig dimethyl oxalate.

Calcium can be determined as the oxalate by precipitation from homogeneous
solution by cation release from the EDT.A complex in the presence of o.xalatc ion
(Ref. 3).

(</) Sulpliuicw Sulphate ion may be uenerated bv the hydrolysis of
sulphamicacid;

NiUSOjH + 1 KO = Nil., * -f H * + SOr '

The reliction has been used to produce barium sulphate in a coarsely crystalline
form.

1 he hydrolysis of dimethyl sulphate also provides a .source of sulphate ion, and

412

GRAVIMETRY XI, 8

the reaction has been used for the precipitation of barium, strontium, and
calcium as well as lead ;

(CH3)2S04 + 2H20 = 2CH3OH+2H+

XI, 8.‘ WASHING OF THE PRECIPITATE. The experimental aspect of this
important subject is dealt with in Section HI, 42. Only some general theoretical
considerations will be given here. Most precipitates are produced in the presence
of one or more soluble compounds; and it is the object of the washing process to
remove these as completely as possible. It is evident that only surface impurities
will be removed in this way. The composition of the wash solution will depend
upon the solubility and chemical properties of the precipitate and upon its
tendency to undergo peptisation, the impurities to be removed, and the influence
of traces of the wash liquid upon the subsequent treatment of the precipitate
before weighing. Pure water cannot, in general, be employed owing to the
possibility of producing partial peptisation of the precipitate and, in many cases,
the occurrence of small losses as a consequence of the slight solubility of the
precipitate: a solution of some electrolyte is employed. This should possess a
common ion with the precipitate in order to reduce solubility errors, and should
easily be volatilised in the preparation of the precipitate for weighing. For these
reasons, ammonium salts, ammonia solution, and dilute acids are commonly
employed. If the filtrate is required in a subsequent determination, the selection is
limited to substances which will not interfere in the sequel. Also hydrolysable
substances will liecessitate the use of solutions containing an electrolyte which
will depress the hydrolysis (compare Section H, 18). Whether the wash liquid' is
employed hot or at some other temperature will depend primarily upon the
solubility of the precipitate; if permissible, hot solutions are to be preferred
because of the greater solubility of the foreign substances and the increased speed
of filtration.

It is convenient to divide wash solutions into three classes :

1. Solutions which prevent the precipitate from becoming colloidal and passing
through the filter. This tendency is frequently observed with gelatinous or
flocculated precipitates but rarely with well-defined crystalline precipitates. The
wash solution should contain an electrolyte. The nature of the electrolyte is
immaterial, provided it is without action upon the precipitate either during
washing or ignition. Ammonium salts are therefore widely used. Thus dilute
ammonium nitrate solution is employed for washing iron(III) hydroxide
(hydrated iron(III) oxide), and 1 per cent nitric acid for washing silver chloride.

2. Solutions which reduce the solubility of the precipitate. The wash solution
may contain a moderate concentration of a compound with one ion in common
with the precipitate, use being made of the fact that substances tend to be less
soluble in the presence of a slight excess of a common ion. Most salts are
insoluble in ethanol and similar solvents, so that organic solvents can sometimes
be used for washing precipitates. Sometimes a mixture of an organic solvent (e.g.,
ethanol) and water or a dilute electrolyte is effective in reducing the solubility to
negligible proportions. Thus 100 cm^ of water at 25 °C will dissolve 0.7 mg of
calcium oxalate, but the same volume of dilute aimnonium oxalate solution
dissolves only a negligible weight of the salt. Also lOOcm^ of water at room
temperature will dissolve 4.2 mg of lead sulphate, but dilute sulphuric acid or 50
per cent aqueous ethanol has practically no solvent action on the compound.

3. Sohaiotis which prevent the hydrolysis of salts of weak acids and bases. If the

413

GRAVIMETRY XI, 10

(d) essential water, present as water of hydration or crystallisation (e.g.,
CaC204,H20 or Mg(NH4)P04.,6H20) or as water of constitution (the water
is not present as such but is formed on heating, e.g.,
Ca(0H)2->Ca0 + H20).

In addition to the evolution of water, the ignition of precipitates often results
in thermal decomposition reactions involving the dissociation of salts into acidic
and basic components, e.g., the decomposition of carbonates and sulphates; the
decomposition temperatures will obviously be related to the thermal stabilities.

The temperatures at which precipitates may be dried or ignited are determined
from a knowledge of the thermogravimetric curves for the individual substances
(the nature of these curves is explained in detail in Chapter XXIII). Thus calcium
oxalate remains as the anhydrous salt between about 200° and 350 °C, above
which it progressively decomposes. By using these curves it is possible to select
appropriate drying and ignition temperatures to convert precipitates into pre-
determined chemical forms. Thermogravimetric curves must be interpreted with
due regard to the fact that in obtaining them the temperature is changing (usually
at a regular rate), whereas in routine gravimetric analysis a precipitate is brought
to a specified temperature and maintained at that temperature for a definite time.

Quantitative separations based upon precipitation
methods

XI, 10. FRACTIONAL PRECIPITATION. The simple theory of fractional
precipitation has been given in Section II, 10. It was shown that when the
solubility products of two sparingly soluble salts having an ion in common differ
sufficiently, then one salt will precipitate almost completely before the other
commenees to separate. This separation is actually possible for a mixture of
chloride and iodide, but in other cases the theoretical predictions must be verified
experimentally because of the danger of co-precipitation (Section XI, 5) affecting
the results. Some separations based upon fractional precipitation, which are of
practical importance, will now be considered.

A. Precipitation of sulphides. In order to understand fully the separations
dependent upon the sulphide ion, we shall consider first the quantitative
relationships involved in a "^saturated solution of hydrogen sulphide. The
following equilibria are present :

H2S^H+-(-HS-

HS-^H+-I-S^-

[H + ]x[HS-]/[H2S] = Xi = 1.0x10-’ (1)

[H+]x[S 2-]/[HS-] = X2 = 1.0x10-’'^ (2)

The very small value of K 2 indicates that the secondary dissociation and
consequently [S^"] is exceedingly small. It follows therefore that only the
primary ionisation is of importance, and [H ] and [HS “ ] are practically equal in
value. A saturated aqueous solution of hydrogen sulphide at 25 °C, at

415

XJ, 10 QUANTITATIVE INORGANIC ANALYSIS

atmospheric pi essiirc, is appro.\inialc)y 0. 1 .l/.anci calculation shows (sec Scxiion
II, 6) that in this solution

[irj - (HS' ] -- I X !0' 'moklm''-’.

(S-'j = 1 X l()~''‘moitlrn''-*.

and {S-'j is inversely proportional to the square of the hydrogen ion
concentration. Clearly, hy varying the pH of the solution the sulphide ion
concentration may K- controlled, and in this way, separation.s of metallic
sulphides may be clfected.

As shown in Section IF. 7. in a solution of 0.25,1/ hydrochloric acid saturated
with hydrogen sulphide (thi-s is the solution employed for the precipitation of the
sulphides of the Group 11 iitetaK in qualitative analysis).

[HS“) - 4.V 10' *11101 dm ' ‘

and

(S* ']=■• 1.6 .< 10' -‘mol dm

Thus by changing the acidity from 0.5 x 10" *5/ lihal pre.sent in .satuniled
hydrogen sulphide water) to 0.25,U. the sulphide ion concentration ic reduced
from 1 s 10 to 1.6 x 10'

With the aid of a table of solubility products of metallic sulphides (sec
Appendi.v), we can calculate whether certain sulphides will precipitate under any
given coiidition.s of acidity and .iKo the concentration of the metallic ions
remaining in solution, 1‘rccipitatiun of .i metallic .sulpliide ,MS uill occur when
IM’ ' ) X (S*' j e.vcccds the sulubitiiy product, »uiii the concentration of metallic
ions remaining in the solution may be calculated from the equation:

IM*’1

“^vtS

IS*-'!

1.0 X 10' •* .s{H,.Sl

(3)

A.S an example vve may consider the precipitation of coppcr(ll) sulphide
(Ksc-s - ^-5 X 10"*-') and ironil!) sulphide 1-5 x 10"*'*) from O.Ol.Vf

solutions of the metallic ions in the presence of 0.25,V/-hydrochIoric aeid.
For eoppef(ll) sulphide, the solubility product i.s readily e.xcceded:
[S* ■ } - L6 ;< 10’ jCu- ‘ j ~ 0.01 and precipitation will occur until

(Cu- ' 1 - 8.5 X m- X [H ’ )%T.O X 10"- ‘ x [IkSj
8.5x U)--*'x(0.25)-il.0x 1-=' xO.l

i.e.. precipiialimi is virtually complete. Wiili iron(II) sulphide, the .solubility
product cannot be exceeded and precipitation will not occur under these
conditions. If, however, the acidity is suUicicnily decreased, and con.scquently
IS^ "] increased, iron(II) sulphide will he precipitated.

’I he case of zinc sulphide is of special interest. Various values arc given in the
literature for its solubility product: the most trustwortliy figures vary between
1 X 10 -■* and 8 x 10'*'’. If we accept the latter figure, then we should expect
precipitation to occur in a, .say, 0.01 d/ solution of zinc ion.s in the pre.sence ol
0.25.U-hydrochloric acid, since the S.P. sliould be exceeded; funhermore, the
residual zinc ion concentration should be 4.7 xlO""* when calculated as
described above. In practice, precipitation doe.s not occur at this acidity. This

416

GRAVIMETRY XI, 10

may be partly due to the great tendency that zinc sulphide possesses to remain in
supersaturated solution, but is perhaps best explained as follows. The above
figure for the solubility product refers to a solution in equilibrium with relatively
large particles, whereas for precipitation to occur it is necessary that the S.P. of
the particles actually formed should be exceeded. It may well be that under the
above experimental conditions these are extremely small, thus possessing a
greater solubility (Section XI, 3) and a greater solubility product; precipitation
will therefore not take place. This view is supported by the fact that post-
precipitation of zinc sulphide (compare Section XI, 5) will occur upon the surface
of other metallic sulphides, such as those of copper and mercury. It is possible to
precipitate zinc in acid solution provided the experimental conditions are very
carefully controlled, e.g., when the pH of the solution lies between 2 and 3 and
ammonium salts are present as coagulants : this is attained by the use of a buffer
mixture of formic acid and ammonium formate, sulphuric acid and ammonium
sulphate or of chloroacetic acid and sodium chloroacetate. It is probable that
large particles of zinc sulphide are initially formed under these conditions.

It must be pointed out that the above calculations are approximate only, and
may be regarded merely as illustrations of the calculations involved in
considering the precipitation of sulphides under various experimental
conditions ; the solubility products of most metallic sulphides are not known with
any great accuracy. It is by no means certain that the sulphide ion ~ is the most
important reactant in acidified solutions; it may well be that in many cases the
active precipitant is the hydrogensulphide ion HS”, the concentration of which
is considerable, and that intermediate products are formed. Also much co-
precipitation and post-precipitation occur in sulphide precipitations unless the
experimental conditions are rigorously controlled.

B. Precipitation and separation of hydroxides at controlled hydrogen-ion
concentration or pH. The underlying theory is very similar to that just given for
sulphides. Precipitation will depend largely upon the solubility product of the
metallic hydroxide and the hydroxide-ion concentration, or since
pH+pOH = (Section II, 16), upon the hydrogen-ion concentration of the
solution.

We have seen that the sulphide ion concentration of a saturated aqueous
solution of hydrogen sulphide may be controlled within wide limits by suitably
changing the concentration of hydrogen ions — a common ion — of the solution.
In a like manner the hydroxide ion concentration of a solution of a weak base,
such as aqueous ammonia (Xj = 1.8x10“®), may be regulated by the addition of
a common ion, e.g., ammonium ions in the form of the completely dissociated
ammonium chloride. The magnitude of the effect is best illustrated by means of
an example. In a 0. 1 M-ammonia solution, the degree of dissociation is given
(Section H, 5) approximately by :

a = yXF = 71.8 X 10“® X 10 = 0.0013

Hence [OH-] = 0.0013, [NH^+j = 0.0013, and [NH 3 ] = 0.0987. As shown in
ection n, 7, by the addition of 0.5 mole of ammonimn chloride to 1 dm® of this
solution, [OH ] is reduced to 3.6 X 10“ ® mol dm“®.

Thus the addition of half a mole of ammonium chloride to a 0. 1 M solution of
aqueous ammonia has decreased the hydroxide-ion concentration from 0.00 1 3 to

417

XI, 10 QUANTITATIVE INORGANIC ANALYSIS

0.0000036, or has changed pOM from 2.9 to 5.-1, i.c., the pH has changed from
11.1 to 8.6.

An imnicdiatc application of the use of the aqucou.s ammonia-ammonium
cliloridc mixture may be made to the familiar e.xample of the prevention of
precipitation of magnesium hydro.xide (S.F. 1.5 x 10""). We can first compute
the minimum hydroxide ion concentration necessary to prevent precipitation in,
say, 0. 1 .)/-magncsium solution.

lOH"}

pOH

•1.9

'■[Mg-'n '

and pH

■■'ot"

: l-l,0-4.9

^ 1.22 X

9.1

If sve employ an aqueous ammonia solution which is 0.1,)/', the concentration of
[NIU^ I ion as ammonium chloride or other ammonium salt necessary to prevent
the precipitation of magnesium hydro.xide cun be readily calculated as roliovv.s.
Substituting in the mass-action equation;

INHjf •'
[NH/l x 1.22 X 10

U.l

- •S *” t <- I A .• 1

-•■ 1.0 X 10

(NlU'l - /.-f-Vx //>"•.)/

This corresponds to an ammonium chloride concentration of
1.48 X 10- ‘ X 53.5 7.9gdm“ '.

We will now consider the conditions necessary for the practically complete
precipitation of niagnesuim hydroxide from a 0.1.)/ solution of, say, magnesium
chloride. /\ pOH slightly m excess of 4.9 (i.c., pi I « 9.1 ) might fail to precipitate
the hydroxide owing to supers.miraiion. Let us suppose the hydro.ude ion
concentration is increased ten times, i.c.. to pOH 3.9 or pH 10. 1, then, provided
no supersaturation is present;

(Ms-'

^'sM.IDIIi, _ 1-5 X 10 "
IOH”')*' ■■ (L22x 10'^)-

0.00 1.\/

i.c., the coneeiiiralion of the magnesium ions remaining in .solution is O.OOl.U, or
1 per cent ot the magnesium ions would lemam unprecipitated. If pOH is
changed to 2.9 or pH to ll.l. ii can be shown in a similar way that the
coiKcnlraiion of the magnesium ions left in solution is cu. 1 x 10" ■*.)/. so that the
precipitation error i.s 0.1 per cent, a negligible quantity. We may therelore say
that magnesium is precipitated quantitatively at a pH of 1 1.1.

Our knowledge of the solubility products of metallic hydroxides is. however,
not very precise, so that it is always po.ssible to make exact theoretical
calculations, 'fiic approximate pH values .it which various hydroxides begin to
prex'ipiiatc Iroin dilute, solution arccollccted in Table Xl,4.

rile precipitaied metallic hydro.xides or hydrated oxides are gelatinous in
character, and they lend to be cojitaminated witJi anions by adsorption and
occlusion, and sometimes with basic salts. The values presented inTablc XI, 4
suggest that many separations should be possible by fractional precipitation of

418

GRAVIMETRY XI, 11

Table XI, 4 pH values at which various hydroxides are precipitated

Metal ion

' Metal ion

Sn^'^,Fe^+,Zr'‘+

• 7

■Fe2+

Th*+ .

■ • , 8 '

Co^*,NE+,Cd^+

. Ag^,

Mg^^

the hydroxides. These separations are not always practical owing to high local
concentrations of base when the solution is treated with alkali. Such unequal
concentrations' of base result in regions of high local pH and lead to the
precipitation of more soluble hydroxides, which may be occluded in the desired
precipitate. Slow, or. preferably homogeneous, neutralisation overcomes this
difficulty, and much sharper separations may be achieved.

The common tripositive cations may be separated from many dipositive
cations by the basic acetate or basic benzoate method. These separations are based
upon the fact that the equilibria for the first dissociation of the typical ions are :

[M(H20),]3++H20^

[M(H20),_i(OH)P+ + H30+.(A = 5 X 10-3-1 X lO"")
[M(H20),]3++H20-

[M(H20),_,(OH)r+H30+(i(:= 10-’-10-i2)

Any strong acid that may be present is first neutralised. Then, by selecting an
appropriate base, whose conjugate acid has a of about 10 " 3, the equilibrium
for the trivalent cations will be forced to the right ; the base is too weak, however,
to remove the hydroxonium ions from the equilibrium of the divalent cations.
Since a large excess of the basic ion is added, a basic salt of the trivalent metal
usually precipitates instead of the normal hydroxide. Acetate or benzoate ions (in
the form of the sodium salts) are the most common bases that are employed for
this procedure. The precipitation of basic salts may be combined vrith
homogeneous precipitation, and thus very satisfactory separations may be
obtained.

M, 11. ORGANIC PRECIPITANTS. Separation of one or more inorganic
ions from mixtures may be made with the aid of organic reagents, with which
they yield sparingly soluble and often coloured compounds. These compounds
usually have high molecular weights, so that a small amount of the ions will yield
a relatively large amount of the precipitate. The ideal organic precipitant should
be specific m character, i.e., it should give a precipitate with only one particular
ton. In few cases, however, has this ideal been attained; it is more usual to find
that the organic reagent will react with a group of ions, but frequently by a
rigorous control of the experimental conditions it is possible to precipitate only
one of the ions of the group. Sometimes the precipitated organic compound may.
be weighed after drying at a suitable temperature; in other cases the composition
s not quite definite and the substance is converted by ignition to the oxide of the
metal, in a few instances, a titnmetnc method is employed which utilises the
quantitatively precipitated organic complex.

character and was

directed towards a search for specific, or at least highly selective reaeents for
particular metal ions. A more fundamental approach is now possible, Attention

419

XI, 11 QUANTITATIVEINORGANICANALYSIS

being directed to theoretical factors which lead to selectivity and also to a
quantitative consideration of the equilibria involved. Frequently suflieient
selectiv ity can be achieved for a particular purpose by controlling such variables
as the concentration of the reagent and the pH. and also by taking advantage of
secondary coinplexing agents {nutsking agenls-see Scxnion X, 27),

It is dilliculi to give a rigid classification oflhe iniincrous organic reagents. The
most important arc those which form cheLite complexes, which involve the
formation of one or more (usually five- ors'ix-mcmbered) rings incorporating the
metal ion; ring formation leads to a relatively great stability. One classification of
organic reagents is concerned with the number of hydrogen ions displaced from a
neutral molecule in forming one chelate ring. A guide to qualitative predictions
about the applicability of organic reagents for analytical purposes may be
obtained from a study of the formation constant of the coordination compound
(which is a measure of its stability), the elfect olThe nature of the metallic ion and
of the ligand on the stability of complexes, and of the precipitation equilibria
involved, particularly in the production of uncharged chelates. For further
details, llie reader is referred to Sections X, l‘> -21 and to the books on chelate
compounds listed in the Ihhiiography at the end of tiiis chapter. Selected
examples of precipitation reagents folUsw.

Dimethylgljo.vime. This reagent (1) was iliseovercd by L. Tschugaelf
and was applied" by O. IJrunck for the determination of nickel in steel. It gives a
bright red precipitate (11) of NilC'^HiOjiN',). with nickel salt solutions: pre-
cipitation is usually carried vnd in aminoniacal solution or in a bulTer .solution
containing ammonium .icctatc and acetic acid. The complex is weighed after
drying at 110- 120 C. A slight cxccvS of the re.igent exert.s no action on the
precipitate, but a large excess should be avoided because: (o) of the possible
precipitation of ihcdimciliylglyovime itself due to its low solubility in water (it is
used in ethanolic. solution),* and (h) the increased solubility of the precipitate in
waicr-cihunol mixiures. * The interference of iront 111), aluminium, or bi.smuih is
psevewtesl fey Use ■.vdfewwvw of a >.ofevfek V .wU afe os efer .fee ; wfeew wmefe cofeafe, mi,
or manganese is present, precipitation should t.ikc place in a sodium acetate,
rather than an ammonium .leeiate. buffer.

Solutions of palladiumdl ) salts give a charaeterisiic yellow precipitate in dilute
hydrociiloric or sulpluirie acid solution; the eomposiiion is similar to that of
nickel, viz., PdlC^lFO^N-)-. and the precipitate can be dried at 1 1()-120Cand
weighed. The precipitate is almost insoluble in hot water, but dissolves readily in
ammonia and cy.mide soiiiiions. Gold is reduced to the metal by the reagent, and
platinum (if present in appreciable quantity) is partially precipitated either as a
greenish comple.x compound or as the metal, upon boiling the solution. The
precipitation of palladium Ls not complete in the presence of nitrates.

.Solutionis of bismuth salts in the presence of FDTA give a yellow precipitate
with dimcthylglyo.xime; precipitation is quantitative at pii 11.0-11.5, the
precipitate is believed to be a polymer (Ref. 4) with an apparent composition
13i,Oj(C4l IfaNiOj). and may be dried at 105 125 ’C. In the presence of EDTA
anti of cyanide ion. A). As. Ba. Cd, Ca. Co. Cu, Fb. Mg. Hg, Ni. Pd, Pt, Ag, Sr.
W. and Zn do not interfere.

• These po,s,ihic errors may beaw'iJcJ bycnipIi)ymi;ilisoduinuhiiicili)lgl>ovimc. 'thichissolublcin
Wiiicr—scc below.

420

GRAVIMETRY XI, 11

CH3— C=NOH

CHj— C=NOH

(I)

CH,— C=N

.N=C— CH3

CH3-C=n/ ^ N=C-CH3

''h/

(II)

Dimethylglyoxime is only slightly soluble in water (0.40 g dm ), con-
sequently it is employed as a 1 per cent solution in ethanol. The sodium salt of
dimethylglyoxime Na2C4H602N2,8H20 is available commercially: this is
soluble in water and maybe employed as 2—3 per cent aqueous solution.

Furil-a-dioxime (III) has also been proposed for the determination of nickel. It
gives a red precipitate with nickel salts in ammoniacal solution. The complex is
less soluble than nickel dimethylglyoxime, and has a smaller nickel content, thus

CH3CH C=NOH

1 I

CH2 C=NOH

(IV)

giving a larger weight of precipitate for a given weight of nickel. The great
advantage of furil-a-dioxime is its solubility in water, which precludes the
possibility of contaminating the precipitate of the nickel derivative with the free
reagent. A 2 per cent aqueous solution is normally used. The reagent is, however,
expensive.

Cyclohexane-l,2-dione dioxime (nioxime) is more soluble in water (8.2 g dm“ ^
at 21 °C) than dimethylglyoxime: it is an excellent reagent for the gravimetric
determination of palladium, but an empirical factor is required for the
determination of nickel owing to co-precipitation of the reagnet.
4-Methylcyclohexane-l,2-dione dioxime (4-methyl-nioxime) (IV) is fairly soluble
in water (3.4 g dm“ ^ at 25 °C) and precipitates nickel quantitatively as a scarletr
coloured complex down to pH 3 ; the precipitate is uncontaminated by excess of
reagent and filters easily : it is equally useful for the gravimetric determination of
palladium.

B. Cupferron (ammonium salt of iV-nitroso-M-phenylhydroxylamine), (V).*
This reagent, the ammonium salt of nitrosophenylhydroxylamine, forms
V).* This reagent, the ammonium salt of nitrosophenylhydroxylamine, forms
insoluble compounds with a number of metals in both weakly acid and strongly
acid solutions. It is most useful when employed in strongly acid solutions (5-10
per cent by volume of hydrochloric or sulphuric acid) and then precipitates
iron(III), vanadium(V), titanium(IV), zirconium(IV), cerium(IV), niobium(V),

421

XI, 1 1 QUAHTiT/\Tl VE INORGANIC ANALYSIS

uuUalum(V), tungsica(Vl), galliumdU) and tin(iV),
.separating these elements from aluminium, beryllium,
chromium, mangane.se, nickel cobalt, zinc, uranium(Vl),
calciunt, strontium and barium. The pre,scncc of tartrate
and o.xalate has no ciTeet upon the precipitation of metals
by cuiderron.

Tlic cupferron method is very satisfactory for the separation of iron,
titanium, zirconium, vanadium and. in .special cases, tin, tantalum, uranium,
and gallium.

The reagent is usually employed as a <> per cent aqueous solution ; this should
be freshly prepared, since it does not keep satisfactorily for more than a few days.
The solid reagent should be stored in amber bottles containing a few lumps of
ammonium carbonate. Precipitation is ahv.i>s carried out in the cold, since
cupferron is decomposed into niirosobenzene on healing. Suilkieni reagent is
added to form the curdy piccipitalcoflhc metallic derivative ofcupfcrron and to
give a white ilocculent precipitate of free nitrosophcn>lh>dro.xylamine (needles).
Precipitates slioutd Ise liitered as soon after their formation as possible, since
excess of cupferron is not scry si,ihle in acid solution. Nitric acid solutions
cannot be used for liie precipitation, since oxidising agents destroy the reagent,
The addition of macerated filter p.ipcr assisl.s the lillralion of the precipitate and
also the sub.sequent gradual ignition. The precipitates cannot be weighed after
drying, but must be ignited to the corresponding oxide and weighed in this form.
The Ignition must he done cautiously in a /urgi' crucible w iih a gradual increase in
temperature to avoid meciianica! loss.

Neo-cupferroii(aniraoniumsallof.V-nitroso-jV-2-naphthjbhydro\jlanunc),(VI)
forms less soluble and more bulky piccipitates than cupferron. It may be
employed for the direct separation of iron and copjxT in mineral and sea-walcrs
without preliminars’ concenir.ition.

.N-i}cnzoyl-.V-plieU)lh)dro\)luniiue, Coll.iC0(C\,Hj)N01I, has been pro-
posed as a reagent similar to cupferron in its reactions, but is more stable. The
reagent is moderately soluble in hot water but e.i.sily soluble in ethanol and other
organic sohcnls. The Cu(ll), i'eillU and .A1 compic.xcs can be weighed us such
Ic.g., as Cu(C| jUjoOjN), ;■ but thcTi compound inust be ignited to the o.vide.

C. S-lIydroxyquinoline (oxine), (VU). Oxine (C.jiljON) forms sp.iriiigiy
soluble derivatives with metallic uins. which have the composition M(C'.,iloON)’
if the co-ordination number of the metal is four (e.g.. magnesium, zinc, copper,
cadmium, lead, and indium), M(C,M„ON)j if tlie co-ordination number is six
(e.g., aluminium, iron, bismiuh, and gallium), and XUCjHcON)^ if the co-
ordination number is cigiit (e.g.. thorium and zirconium). There are, however,
.some exceptions, for example. TiO(CoH^ON)-. MnOjfC^HcON),,
WO,(C;f l„ON),. and UO,(C.„HpON),. By proper control of the pH of the
solution, Iry the use of comple.x-forming teagents and by other nicthod.s.
numerous separations may be carried out; tiius aluminium may be separated

• The n.iine cupScrtun u.is .usMiincii to lliecoiiipoumt b> O. Il.iiiJiscli, . iikI iv JcfiM-sl Irom ihc Ijcl thst
Ific resgenl prccipii.ilcs b,)ih copjscr .imt u,)ii Cuplcuoa piectpuaics iron coniplcleiy in sirong
iniiicr.il jcid soliilion. and cop|vr i, ,)nl> i{iun(iutisel) prectpii.itcd in l.mill)' aeid sohilion. The
selcetnily of (he rc.igcnl is picalc’.! in stiunirly aetd sotulioii.

•»22

GRAVIMETRY XI, 11

from beryllium in an ammonium acetate-acetic acid buffer, and magnesium from
the alkaline-earth metals in ammoniacal buffers. The pH values, extracted from
the literature, for the quantitative precipitation of metal oxinates are collected in
Table XI, 5.

Table XI, 5 pH range for precipitation of metal oxinates

Metal

initial

precipitation

complete

precipitation

Aluminium

2.9

4.7- 9.8

Bismuth

3.7

5.2- 9.4

Cadmium

4.5

5.5-13.2

Calcium

6.8

9.2-12.7

Cobalt

3.6

4.9-11.6

Copper

3.0

3.3 +

Iron(III)

2.5

4.1-11.2

Lead

4.8

8.4-12.3

Magnesium

7.0

8.7 +

Manganese

4.3

5.9- 9.5

Molybdenum

2.0

3.6- 7.3

Nickel

3.5

4.6-10.0

Thorium

3.9

4.4- 8.8

Titanium

3.6

4.8- 8.6

Tungsten

3.5

5.0- 5.7

Uranium

3.7

4.9- 9.3

Vanadium

1.4

2.7- 6.1

Zinc

3.3

4.4 +

8-Hydroxyquinoline is an almost colourless, crystalline solid, m.p. 75-76° ; it is
almost insoluble in water. The reagent is prepared for use in either of the
following ways :

(n) Two grams of A.R. oxine are dissolved in lOOcm^ of 2Af-acetic acid, and
ammonia solution is added dropwise until a turbidity begins to form; the
solution is clarified by the addition of a little acetic acid. This solution is stable for
long periods, particularly if it is kept in an amber bottle.

[b) Two grams of A.R. oxine are dissolved in 100 cm^ of methanol or ethanol
(this reagent cannot be used for the determination of aluminium) or in acetone.
The solution is stable for about ten days if protected from light. It is stated that
the alcoholic solution may be employed in cases where precipitation occurs at a
high pH, and the acetic acid solution for precipitations at low pH.

The following general conditions for conducting precipitations with 8-
hydroxyquinoline may be given :

1 . The reagent is added to the cold solution (or frequently at 50-60 °C) until the
yellow or orange-yellow colour of the supernatant liquid indicates that a small
but definite excess is present.

423

XI, 11 QUANTITATIVE INORGANIC ANALYSIS

2. The prccipiuuc is coiigulatcd by a short period of healing at a temperature
not exceeding 70 C.

3. The precipitate may be filtered through paper or any variety of tiltcring
crucible.

4. The filtrate should possess a yellow or orange colour, indicating the
presence of exce.ss of precipitant. If a turbidity appears, a portion should be
healed; if the turbidity disappears, it may be assumed to be due to excess of
reagent crystalli.sing out, and i-s harmless. Otherwise, more reagent should be
added, and the solution filtered again.

5. \Va.shing of the precipitate may often Ik elfected ’.yilh hot or cold water
(according to the solubility of the metal ‘oxinatc') and is continued until the
(iltrates become colourless. The use ofethaiiol is permissible if it i;s known to hate
noelleci upon the precipitate.

6. The washed precipitate may be dried at 105-llO C (Usually hydrated
‘oxinatc') or at 1.10 MO C (anliydrous 'oxin.iie'). In cases vsliere prolonged
hcMting at 130-140 C is required, slight decomposition may CKCur. Frequently
ignition to the oxide yields a more suitable form for weighing, but care mast be
exercised to present loss, since many 'oxinates' arc appreciably volatile; it is
usually best to cover the complex w ith oxalic acid ( 1 -3 g) and heal gradually. The
determination may also be completed tilrimetric.illy by dis.solv ing the precipitate
m dilute hydriKhluric acid and titrating with a standard solution of poias.sitim
bromatc a.s detailed in Section .\, 140.

D. Uen^oia-a•uxi^^c (cuprun). (VI ill. This compound yields a green pre-
cipitate, CuC'i jH, 'topper m dilute ammoniacal sohilion, which may-

be dried to constant wciglit at 1 10 C. Ions which ;irc precipitated by aqueous
ammonia arc kept m solution by the addition of tartrate: the reagent is then
specific for copper. Copper may thus be separated from cadmium, lead, nickel,
cobalt, /inc, aluminium, and small amuum.s ofiron.

From strongly acidic solutions ben/om- r-oxime piccipitales molybdate and
umgstaie iosvs quantuaiixely , ebreunate, vanadate, mobatc, tanial-.uc, and
palladium! 11) are partially precipitated, fhe moly bd;iie complc.x is best ignited at
500'525 ’C to MoO, before weigliing; alternatively, the precipitate may be
dissolved in ammonia solution ;ind the molybdenum precipitated as lead
molybdate, in winch form it is conveniently weighed.

Benzom-x-oximc Ls a while, cfysi.ilhnc solid, m.p. 152 C. which is sparingly
soluble m water but lairly soluble m ethanol. The reagent is employed as a 2 per
cent solution in ethanol.

QH,— cil— on cn=NOH

i-'"/

C„nj— C=NOH '-x-\

(Vlll) (IX)

E. Salicylaldehydeo\inie(lX). Thiscompound ischielly employed tor the
determination ol copper; a greenish-yellow precipitate CidCiHnOiN)' is
obtained in the presence of acetic acid (liie precipitation is complete at pH 2.6),
which is weighed after drying at 100-105 C. Iron(in) i.s carried down with the
copper complex in acetic acid solution and interferes seriously, but silver.

■124

GRAVIMETRY XI, 11

cadmium, mercury, arsenic, and zinc have no effect. Salicylaldehyde oxime reacts
with many other ions, and has found application in the determination of lead,
bismuth, zinc, nickel, and palladium. As with similar non-selective reagents (e.g.,
oxine) the pH of the solution is an important factor, particularly if it is desired to
separate one divalent metal from another. Thus copper is completely pre-
cipitated at a pH of 2.6, nickel commences to precipitate at a pH of 3.3, and hence
for the separation of copper from nickel the pH must be maintained between 2.6
and 3.3.

A bright yellow, insoluble basic salt is formed with bismuth ions in almost
neutral solution, which must be ignited to the oxide Bi203 for weighing. Lead is
quantitatively precipitated as a yellow complex PbC7H502N at pH 8.9 or
higher; the use of a strongly ammoniacal solution permits the separation of lead
from silver, cadmium, and zinc. Palladium(II) is precipitated quantitatively as
yellow Pd(C7H602N)2 from acid solution, and can thus be separated from
platinum. Nickel may also be satisfactorily determined as the green complex
Ni(C7H602N)2.

Salicylaldehyde oxime is a white, crystalline solid, m.p. 57 °C, which is
sparingly soluble in water. The reagent is prepared by dissolving 1.0 g of
salicylaldehyde oxime in 5 cm^ of 95 per cent ethanol, and pouring the solution
slowly into 95 cm^ of water at a temperature not exceeding 80 °C; the mixture is
shaken until clear and filtered, if necessary. Another procedure is to add 2.22 g of
pure salicylaldehyde dissolved in 8 cm^ of 90% ethanol to 1.27 g of A.R.
hydroxylammonium chloride dissolved in 2 cm^ of water: the resulting solution
is diluted with 15 cm^ of 90% ethanol and poured slowly and with stirring into
225 cm^ of water at 80 °C; when cold, the solution is filtered if necessary and
stored in an amber bottle. The reagent decomposes in solution, and should not be
kept for more than about three days.

F. l-Nitroso-2-napthol (X). This organic reagent precipitates quanti-
tatively cobalt, iron(III), palladium, and zirconium from slightly acid solutions ;
it precipitates partially tin, silver, bismuth, chromium(III), titanium,
tungsten(VI), uranium(VI), and vanadium(V). The following elements are not
precipitated : lead, cadmium, mercury, arsenic, antimony, beryllium, aluminium,
nickel, manganese, zinc, calcium, and magnesium. The principal use of 1-nitroso-
2-naphthol is in the separation of cobalt from large amounts of nickel after any

CH(0H)C02H

iron(III) present has been removed. The red-brown, bulky precipitate obtained
in dilute hydrochloric acid solution is reported to have the' composition
Co(CioHg02N)3, but it is doubtful whether the complex is pure ; careful ignition
in the presence of oxalic acid gives a cobalt oxide to which the formula C03O4 has
been assigned, but this also is not perfectly pure, and should not be used except
when dealing with minute amounts of cobalt. For larger amounts the cobalt
oxide may (a) be reduced in hydrogen in a Rose crueible and weighed as the
metal, or {b) be treated with a few drops of concentrated nitric acid to convert it

425

XI, II QUANTITATIVE INORGANIC ANALYSIS

into the nitrate, the excess of nitric acid expelled by evaporating cautiously, and
then converted into the sulphate by at least two evaporations with concentrated
sulphuric ticid, followed by ;i lew drops of water, tind weighed as C0SO4 after
heating for a short time at -150 oOO C (very' dull red heat); the cobalt sulphate
solution may also be electrolysed and the resulting metal weighed.

l-Nitroso-2-naphthol is a brown powder, m.p. 109 “C; it is insoluble in water.
The reagent is prepared by dissolving 4 g of l-nitroso-2-naphtho] in lOOcni^ of
glacial acetic acid and then adding lOOcin' of hot distilled water. The cold,
filtered solution should be used itmitediaiely.

G. 4-Bromotuandelicucid (XI). Mandclk aeid, C(,ll5CH(Ofl)COOH,isa
highly selective and sensitive reagent for /.irconium: precipitation is usually
clfected in a hydrochloric acid mediuiu. Owing to the high concentration of
nv.mdeiic aeid that mvisi be used and the diUlcully of removing the excess by
washing (unless a special wasit .solution is used -a hot solution containing 2 per
cent HCl and 5 per cent m.indelic acid), because of the appreciable solubility of
zirconium mandelate. 4-hrumom;indcIic acitl is preferred .is a reagent for the
gravimetric determination of /’ireonium: the precipitate i.s insoluble in water and
can be washed with water without loss. The precipitate is ignited to and weighed
as the o.vide.

Precipitation is best clfected in a hydrochloric acid medium (up to 3-'5.1f)and
from hot .solution. Cutll). Cd(ll). lied), lig(ll). Sn(II), Th(lV), Sb(ni) and
Fe(ni) interfere, as do chromate and vanadate. Reduction of chromate, vanadate,
and iron(in) eliminates the interference. The following elentent.s do not interfere:
Be. Mg, Ca, Ba. Zn. Al. Ti(I V'l. V(1 V), Q(IUK .Mn(Ul. Kedl). Co ami Nt.

4-Bromomatidclic acid is .1 white ery,sial!ine solid, m.p. I IS', and is slightly
soluble in water, 'llie reagent for prccipiiaiion is a O.l.U-solution; it remains
stable mdetinitely.

II. Nitron (XU). The strong organic base 4. 5-dihydro- l,-Tdiphenyl-.3, 5-
phenylimino-l,2,4-iri.izo!e, which is named nitron, yields a sparingly soluble
cr>.slai!ine nilrale C,„11 ,„n/HNO, in solutions acidi-
C„Uj N C„H5 lied with acetic or sulphuric acid. Perchlorate, per-
1 ! rhenatc. tetralluoborate. and tungstate also form

N C insoluble salts, and can be determined in a similar

m.inner. Numerous other anions, including bromide,
iodide, ehloraie. thiocyanate, nitrile, and chromate,
interfere, but may easily be removed by preliminary
ireaimenl. The results m the presence of chloride are
generally high, possibly because of co-precipitation.

Nitron is a yeliovv, cry stalline solid, m.p, 1S9 C, w hich
is insoluble in water. The reagent consists of a 10 per cent
solution in 5 per cent acetic acid: it slioukl be filtered, if
nece.s.sary. and tlic clear solution protected from light.

I. I aiinic acid. This reagent is e.ssentially a colloidal suspension of
negatively charged parlicle.s capable of tloccuhiling positively charged
hydrated oxide sols, such as tho.se of \VO,.Nb,Oj, and Ta.Oj. Thas if a
tungstate solution is treated with tannic acid ;ind acidilied most ol the tungsten is
prceipnaled: a small amount remams colloidally dfspersed and imiy be
flocculated with a tannic acid precipitant such as cinchonine. The separation of
various elements depends to ;t large extent upon the proper adjustment of the pH
of the .solution.

CH N
N'

QH,

(Xll)

426

GRAVIMETRY XI, 11

This reagent in the form of a freshly prepared 3 or 1 0 per cent aqueous solution
is useful for the separation of some of the so-called rarer elements. It may be
employed inter alia for the quantitative determination of titanium and tung-
states, for the separation of aluminium, chromium, iron, etc., from beryllium,
and of niobium from tantalum (the latter is precipitated selectively from a
slightly acidic oxalate solution). In most cases the flocculent precipitate of the
tannic acid complex of the element is ultimately ignited and weighed as the oxide.

J. The arsonic acids, R-AsO(OH) 2 . Alkyi and arylarsonic acids are of
particular value for the precipitation of the quadrivalent metals (tin, thorium,
titanium, and zirconium). Phenyl- (R = CgHs), propyl- (R = C 3 H 7 ), and 4-
hydroxyphenyl- (R = HO-CgHJ arsonic acids are available commercially; the
last-named is the least expensive.

Phenylarsonic acid, employed as a 10 per cent’ aqueous solution, will
precipitate tin from fairly concentrated acid solutions, and separates it from all
the common elements except titanium and zirconium. Thorium is precipitated
quantitatively from acetic acid-ammonium acetate solution; rare earths and
aluminium do not interfere, but titanium, zirconium, hafnium, and other
quadrivalent ions are precipitated. The reagent provides a good separation of
thorium from the rare-earth elements. The thorium phenylarsonate is dissolved
in dilute hydrochloric acid, precipitated as the oxalate, ignited, and weighed as
ThOj. ’ ^

Propylarsonic acid, employed as a 2.5 per cent aqueous solution, precipitates
zirconium but not titanium in strongly acid solution. The following elements do
not interfere: aluminium, chromium, cobalt, nickel, copper, vanadium, uran-
ium, thorium, and molybdenum; a possible exception is tin, which can be
removed by heating the ignited oxide with ammonium iodide. The precipitate
may be ignited to the oxide; heating is carried out first with a Bunsen and then
with a Meker or Fisher burner.

4-Hydroxyphenylarsonic acid, employed as a 4 per cent aqueous solution,
gives precipitates with titanium and with zirconium in acid solution, and permits
the separation of these elements from iron and all the common elements except
tin and cerium(IV). Hydrogen peroxide prevents the precipitation of titanium,
but does not affect the quantitative separation of zirconium.

K. Pyridine. Pyridine forms insoluble complexes with the thiocyanates of
cadmium, copper, nickel, cobalt, zinc, and manganese; these have the general
formula M(SCN) 2 .(C 5 H 5 N)„ (n= 4 for Co, Ni, and Mn; n= 2 for Cu, Cd, and
Zn). In practice, alkali thiocyanate and a few cm^ of pure pyridine are added to
the neutral or very faintly acid solution of the metal ions. The complexes are
readily filtered. They are washed first with water, then with dilute ethanol (both
containing a little alkali thiocyanate and pyridine), followed successively by
absolute ethanol and diethyl ether, each containing a little pyridine. The
precipitates are weighed after drying in a vacuum desiccator for 5-30 minutes at
the laboratory temperature. The’ method is rapid, but, as is evident, many ions
interfere. The results for manganese are not satisfactory because of the slight
solubility of the complex in the wash solutions. The method is applicable in the
presence of alkali and alkaline-earth metals and of magnesium; considerable
quantities of ammonium salts must be absent.

L. Anthranilic acid (XIII). The sodium salt of anthranilic acid precipitates
in neutral or weakly acid solution zinc, cadmium, cobalt, nickel, copper, lead
silver, and mercury. Several of these salts, including the anthranilates of

427 '

XI, 1 1 QUANTITATIVI- INORGANIC ANALYSIS

into tiic uitnitc, the excess of tutric ucicl expelled cxupoiuting Cviuiiously^ und
then converted into the sulphate by at le'asl two evaporations with concentrated
sulphuric acid, iollotvcd by a Jew* drops ofvvMter, and weighed us CoSO^ after
licittirig for a short time at diO-SyO C (very dull red heal); the cobalt sulphate
solution luay also be electrolysed and the resulting tnelal W'cighcd,

i-Nitro.so-2-naphihoI is a brown powder, m.p. iO‘i 'C; it is insoluble in water.
The reagent i.s prepared by dissolving 4g of l-nitroso-2-naphtliol in lOOcm^of
glacial acetic' acid and then adding lOOenv' of hot dislilted water. The cold,

filtered solution should be used iinniedialeiy.

G. -l-Bromoniandelicaeii] (XI). Mandelic acid. Ct,If 5 CH{OH)COOH,isa
highly selective and sensitive* leagenl for zirconium: precipitation Is usually
effected in a hydrochloric acid medium. Owing to the iiigli concentration of
mandelic acid that must be used and the diliiculiy of rennrving the- excess by
washing (unless a special wM.sIi solution is used--a Itoi .solution containing 2 per
ceiii llCl and 5 jx'r cent mandelic acid), because of the apprex’iable solubility of
zirconium maiuielatc, 4-bioinomandeiic acid is preferred as a reagent for the
gravimetric determination of zitcoiuum; liie pre'cipitalc is insoluble in water and
can be w a.slicd with water wiihont loss. The precipitate is ignited to and weighed
as tlie oxide.

Precipilatio:t is best effected in a hydreuihluric acid medium (up to .1-5.1/) and
from hot solution, (.-udl). Cddli. HgdL dg(H). Sndl), Th(lV), Sb(IU) and
Fcdll) intetfcie, as do chrum.itc and vanadate. Reduction of chromate, vanadate,
and irordlHleliimnaies the mierforeiicc. The follow ingcIemenLs do not interfere;
Be. Mg. Cu. Ba. Zn. Ai. Ti(l V;, V'd Vj. Odlll. .MnlU). FcdD. Co and Ni.

4'Bromoraandehc acid i.s a white crystalline solid, m.p. 1 IH'. and is slightly
soluble in water, fhe reagent for precipitation is a t), l,\/.soluiion; it remains
stable indennitely,

n. Nitron (,\ll). Ihe strong organic b.isc 4,5-dili>dfo-l,4-diphenyl-3,5-
phenyrmiino-l,2,4.tria/o!e, which i.s named nitron, yields a sparingly soluble
crystalline nitrate C,.^,H,„N.iUNOi in solutions addi-
C„Hj Qlls f'vd will) acetic or sulphuric acid. Perchlorate, per-
cheiuiie, icir,itluobot.ue, and tungstate also form
insoluble salts, and can be determined m a similar
manner. Numerous other anions, including bromide,
iodide, chlorate, thiocyanate, nitrite, and chromate,
interfere, but may easily l>e removed by preliminary
treatment The rc.sulis in the presence of chloride are
generally high, possibly because ufcu-prccipitation.

Nitron Is a yellow, crystalline solid, m.p. 189 C.whidi
is insoluble in water, flic reagent consdtsufa lOperornt
solution in 5 per cent acetic acid: it should be filtered. if
neccs.sary. and the clear .solution protected from light.

I. Tannic add. Tliis reagent is ewscniially a colloidal suspension of
negatively charged particles capable of lioeciilating positively charged
hydrated o.xide sols, such as those of \V 0 ^.Nb> 05 . -'"vi Ta.Oj. Thus if a
tungstate solution is treated with lunmc acid and ac]diiied must of the tungsten is
precipitated; a small amount remains colloidally dispers,ed and may be
nocculatcvi with a tannic acid precipitant such a^ cinchonine. The separation of
various elcmenis depends to a large c.xicnt upon the proper adjustment ol the pH
of the solution.

N C

! II
ca N
■^/

c„a.

(XH)

GRAVIMETRY XI, 11

This reagent in the form of a freshly prepared 3 or 10 per cent aqueous solution
is useful for the separation of some of the so-called rarer elements. It may be
employed inter alia for the quantitative determination of titanium and tung-
states, for the separation of aluminium, chromium, iron, etc., from beryllium,
and of niobium from tantalum (the latter is precipitated selectively from a
slightly acidic oxalate solution). In most cases the flocculent precipitate of the
tannic acid complex of the element is ultimately ignited and weighed as the oxide.

J. The arsonic acids, R-AsO(OH) 2 . Alkyi and arylarsonic acids are of
particular value for the precipitation of the quadrivalent metals (tin, thorium,
titanium, and zirconium). Phenyl- (R = QHs), propyl- (R = C 3 H 7 ), and 4-
hydroxyphenyl- (R = HO-C 6 H 4 ) arsonic acids are available commercially; the
last-named is the least expensive.

Phenylarsonic acid, employed as a 10 per cent aqueous solution, will
precipitate tin from fairly concentrated acid solutions, and separates it from all
the common elements except titanium and zirconium. Thorium is precipitated
quantitatively from acetic acid-ammonium acetate solution; rare earths and
aluminium do not interfere, but titanium, zirconium, hafnium, and' other
quadrivalent ions are precipitated. The reagent provides a good separation of
thorium from the rare-earth elements. The thorium phenylarsonate is dissolved
in dilute hydrochloric acid, precipitated as the oxalate, ignited, and weighed as
ThOj.

K. Pyridine. Pyridine forms insoluble complexes with the thiocyanates of
cadmium, copper, nickel, cobalt, zinc, and manganese; these have the general
formula M(SCN) 2 .(C 5 H 5 N)„ (n= 4 for Co, Ni, and Mn; n= 2 for Cu, Cd, and
Zn). In practice, alkali thiocyanate and a few cm^ of pure pyridine are added to
the neutral or very faintly acid solution of the metal ions. The complexes are
readily filtered. They are washed first with water, then with dilute ethanol (both
containing a little alkali thiocyanate and pyridine), followed successively by
absolute ethanol and diethyl ether, each containing a little pyridine. The
precipitates are weighed after drying in a vacuum desiccator for 5-30 minutes at
the laboratory temperature. The- method is rapid, but, as is evident, many ions
interfere. The results for manganese are not satisfactory because of the slight
solubility of the complex in the wash solutions. The method is applicable in the
presence of alkali and alkaline-earth metals and of magnesium; considerable
quantities of ammonium salts must be absent.

L. Anthranilic acid (XIII), The sodium salt of anthranilic acid precipitates
m neutral or weakly acid solution zinc, cadmium, cobalt, nickel, copper, lead,
silver, and mercury. Several of these salts, including the anthranilates of

427

XI 1 1 QUANTITATI VK INORGANIC ANALYSIS

catlmiuiu, zinc, nickel, coball, and copper, arc suiiablc for the quamitatiw
precipitation and grav iniclric dolcrniinaliun of these elements ; t he salts have the

{XIII)

'■■I

-N-

-C(X)H

(XIV)

general formula M(Ci}l,,(>:N);, and may be dried at 105-1 lO' C. The pre-
cipitations must be carried out at a controlled pl i range: in too strongly acidic
solutions the precipitates will not form, while in too strongly basic solutions the
organo-meiallic complexes undergo decomposition. .'Xt present, sodium anthra-
nilate is limited in use to tiie precipitation of .i single listed cation front a relatisely
pure solution in which sm.ill atnounts of amnwtnium. alkaline earths, and aikali-
melal salts may be present.

The reagent consists of a 3 per cent aipjcou.s .solution of pure sodium
anihranilatc.

.M. Quinaldic acid (XIV). This organic reagent gives insoluble compic.xes
with capper, cadmium, zinc, manganese, silver. Csibalt, nickel, lead, mercury,
irondl), palladutmdil, and platinunUn), and insoluble basic salts with iron{!iI)
aluminiunt, chromium, beryllium, and titanium. The formation of insoluble
iluinaldates is inlluenced by tlte pH of tlte .solution. Thms copper quinaldate
Cu(C,uH(,NO->j.H;0 {.ifler drying at 110-115 C) may tve precipitated from
relatively acidic so!utiun>. whilst under the .same conditions the more soluble
cadmium and zinc quinaldatcs remain in solution. Complc.ving reagents may
also assist in rendering the reagent more selective, The quinaldatcs of copper,
cadmium, and /me arc well-dcfmcd crvstalhnc sails, which arc readily fiitcred,
vvu.siied, and dried.

The reagent consists of a 2 per cent aqueous solution of ilie acid or its sodium
salt.

N, I’yrogallol (,XVT. Tiii.s compound yields insoluble eompic.x salts with
bismuth and wjth antimonv, and may be empluved for the quantitative
determination o! the.se eleincnis cither alone or m the ptesence of arsenic, lead,
cadmium, or /me.

Oil Pytogallol IS a white solid, m.p, 133-1.34 C, and is

freely soluble in water. 'I'he reagent consists of a 3 per
v-GIi iicoxygenaicd water: alternatively,

,.VQ|] (XV) iXrv’Sallol may be added to the solution for

analysis.

O. Etliyleiiediainine (1.2-diaminoetliai)c) (NihClljCHeNTl,). Ethyl-
enediamine yields a comple.v cation with copper{II) ions:

Cu’* +2NH, Cn.-CU,-Nll- iCu{N{K CH--Cll--NHd-|

S(Cuen.]-'

This reacts with the eompic.x ions or (Cdl^l'" to yield the insoluble

complex salts (Cu enJ|Hg!.;| and [Cuen,)(Cdl J respectively:

HgCU f4Kl =- K,lHuUl-b2K.Ci
K2[Ifgl4l + lCucn.!(iN(3j): = [Cucn-HUgl^l-f 2KNOi

428

GRAVIMETRY XI, 11

The complex salts are insoluble in water, 95 per cent ethanol, and diethyl ether,
and hence may be employed in the rapid determination of mercury, cadmium,
and copper respectively. The mercury complex is stable in air and in a vacuum,
and its precipitation is unalfected by the presence of ammonium salts ; a valuable
rapid method is thus available for the determination of mercury. The cadmium
complex has similar properties, but is slightly soluble in the presence of
ammonium salts or in strongly ammoniacal solution.

. The reagent may be prepared by either of the following methods :

(a) Heat an aqueous solution containing 1 part of copper(II) nitrate and 2
parts of ethylenediamine on a water bath until a crust forms on the surface of the
violet-blue solution. Allow to cool, filter off the separated crystals of
[Cuen2](N03)2,2H20 at the pump, and wash them several times with ethanol,
followed by diethyl ether. A concentrated solution of this salt is used for
precipitations.

(b) Treat a solution of copper(II) sulphate with an aqueous solution of
ethylenediamine (five to six times the theoretical quantity) until the dark blue-
violet coloration, due to the [Cuen2]^''' ion, appears and does not increase in
intensity upon further addition of ethylenediamine. The presence of excess of
the latter in the reagent has no harmful influence. Here the reagent consists of
a solution of [Cu en2]S04., and is as satisfactory as (a) for the determination
of mercury.

P. 8-Hydroxyquinaldine (XVI). The reactions of 8-hydroxyquinaldine
are, in general, similar to 8-hydroxyquinoline (C), but unlike the latter it does not
produce an insoluble complex with aluminium. In acetic acid-acetate solution
precipitates are formed with bismuth, cadmium, copper, iron(II) and iron(III),-
chromium, manganese, nickel, silver, zinc, titanium (TiO^"*"), molybdate,

tungstate, and vanadate. The same ions are precipi-
tated in ammoniacal solution with the exception of
molybdate, tungstate, and vanadate, but with the
I addition of lead, calcium, strontium, and magnesium;

OH (XVI) aluminium is not precipitated, but tartrate must be
added to prevent the separation of aluminium
hydroxide. ' ’

8-Hydroxyquinaldine (2-methyl-oxine) is a pale yellow, crystalline solid, m.p.
72 °C; it is insoluble in water, but readily soluble in hot ethanol, benzene, and
diethyl ether. The reagent is prepared by dissolving 5 g of 8-hydroxyquinaldine in
12 g of glacial acetic acid and dilutingto 100 cm^ with water: the solution is stable
for about a week.

Q. Tetraphenylarsonium chloride, [(C6H5)4As]^Cl”. This reagent has
been proposed as a precipitant for thallium(IIl) as [(C6H5)4As]'''TlCl4“, which
is weighed in this form. Precipitation is effected in ca. lAT-HCl solution, the
precipitate is washed with lAf-HCl, and then dried at 1 10 °C. Cations that form
insoluble chlorides interfere as do various anions (Br“, I“, F~, N03“, SCN~,
etc.) other than chloride.

R. Sodium tetraphenylboron, Na’^[B(C6H5)4] . This is an excellent re-
agent for potassium: the solubility product of the potassium salt is 2.25 x 10“®.
Precipitation is usually effected at pH 2 or at pH 6.5 in the presence of EDTA.
Rubidium and cesium interfere ; ammonium ion forms a slightly soluble salt and
can be removed by ignition; mercury (II) interferes in acid solution but does not
do so at pH 6.5 in the presence of EDTA. .

429

XI, 12 QUANTlTA nVE INORGANIC ANALYSIS

Xi, 12. VOLATIIJSATION OR EVOLUriON ME I HODS. Evolution or
vohitilisutiori methods depend CN.%eiuial!y upon the removal of volatile con-
stituents. This may be eirecicd in several ways; (i) by simple ignition in air orin a
current of an indilfcrem gas; (ii) by treatment with some chemical reagent
whereby the desired constituent is rendered volatile; and (iii) by treatment witha
chemical reagent whereby the desired constituent is rendered non-volatile. The
volatilised sub.stance may be absorbed in a weighed quantity of a suitable
medium when the estimation is a liin ci one, or the weight of the residue
remaining after the volatilisation of a component is determined, and the
proportion of the constituent calcuhilcd from the loss in weight ; the latter is the
iiitlinxt iiii’tlunl. E.vamplcs of each of titese procedures are given m the following
paragraphs ; full e.vperimerital iletails will be found later in this Chapter.

The determination of snpcrficially bourn! moisture or f)f water of cryslalli-
saliun in hydrated compounds may l>e carried out simply by healing the
substance to a suitable temperature and weighing the residue (see Section .XV'H,
17 for a metliod involving the use of the Karl Eischcr reagent). Substances that
decompose upon heating can be studied mote fully by thermal analysi.s (Chapter
X.XIJI). The water may also be absorbed in .t weighed quantity of an appropriate
drying agent, such a.s anh>drous calcium chloride or magnesium perchlorate.

The determination of carbon dio.xide in carbonate-containing materiaLs may
be effected by treating the sample with excess of acid and absorbing the carbon
dioxide in an alkaline absorbent, such as soda lime, soda lime-asbestos, or
sodium liydro.vide-a.sbe.stos ('ascariic'). fhe g.is i.s completely expelled by heating
the solution and by passing a currem of piirilicd air through the apparatus; it is,
of course, fed tluough a drying agent to remove water vapour before passing to
the earbon•d^oxide-ab^orp^lon apparatus. The gain in weight of tlie latter isdue
to carbon dioxide.

In the determination of carbon m steels and alloys, the substance Ls burnt in
pure oxygen in the presence of caialy Sts. ind tiiccarbon dioxide absorbed asin the
previous example. Prceaiilions arc taken lo remove other volatile constituents
such a.s sulphur dioxide. This method is employed m the determination of carbon
and hydrogen m organic compounds ; the s.jinp!e is burnt in a eonlrolled stream
of oxygen, and tlie water and carbon dioxide arc absorbed separately in .in
appropriate absorbent, c.g . in calcium chloride .saturated with carbon dioxide
and m soda lime (or soda asbestos).

Some clcmeni.s, such as .sodium .ind pota.ssium, in combination vvith radicals of
volatile acids or organic acids, may be determined by evaporating to dry ness with
sulphuric acid; the residual .sulphate is then weighed:

2NaX -f 1 11 SO 4 =■-■ Na^SO^ r 21 iX

Interfering metals must, of eom-sc, be removed first.

.-\n example of a related kind is the determination of pure silica in an impute
ignited silica residue. The latter is treated in a platinum crucible with a mixture
of sulphuric and hydroiliioric acids; the silica is coiueftcd into the volatile
silicon tetralliioride:

SiC),-fqHF“.SiF4i +2H,0

The residue consists of the impurilic.s. and the loss in weight of the crucible gives
the amouni of pure .silica present, provided that thecontaminants are in tliesame
form before and after the hydrofluoric acid treatment and are nut volatilised in

4.10

GRAVIMETRY XI, 13/14

the operation. Although silicon is not the only element that forms a. volatile
fluoride, it is by far the 'most abundant and most often encountered element;
consequently the volatilisation method of separation is generally satisfactory.

The separation of chromium as chromyl chloride Cr02Cl2 is a convenient
method for removing chromium where aluminium and other trivalent elements
are to be determined.

Distillation may be used to separate certain inorganic chlorides and brornides.
Only molecular compounds can be distilled, and so only those elements that can
be converted into volatile halides under certain experimental conditions can be
separated by distillation. Thus arsenic(III), antimony(III), and tin(IV) form
volatile chlorides; only arsenic(III) can be distilled quantitatively from con-
centrated hydrochloric acid. By increasing the b.p. by the addition of phosphoric
acid, antimony (III) can be separated. Tin can be removed by distillation from a
mixture of concentrated hydrochloric and hydrobromic acids.

Boron in the form of borates or boric acid can be separated from complex
mixtures by distilling in acid solution with methanol. The boron volatilises as
methyl borate, is collected in water or other suitable reagent, and is determined
by titration, after the addition of mannitol, with standard alkali.

Practical gravimetric analysis

XI, 13. GENERAL DISCUSSION. Before commencing experimental work
in gravimetric analysis, the student should be familiar with the general theory
underlying the chief experimental processes outlined in Sections XI, 1-12. He
should also read the account of' the technique of gravimetric analysis given in
Sections HI, 33-45, this will assume a greater significance when the various
processes have actually been employed in practice. It is proposed, in the first
place, to give an account of a number of typical gravimetric determinations.
These determinations may be performed with substances which are readily
obtainable in a state of purity (e.g., of analytical reagent quality), and the
experimental error can therefore be checked by calculation. Many may, however,
prefer to carry out the analyses with solutions or solids of ‘unknown’
composition. These determinations should be carried out before those described
under the heading of ‘Systematic gravimetric analysis’ are attempted. In general,
the experimental procedures will not be given in such detail in this latter section
which is arranged in alphabetical order with cations (metals) first, followed by
anions.

For all gravimetric determinations described in this chapter, the phrase ‘Allow
to cool in a desiccator’ should be interpreted as cooling the crucible, etc., provided
with a well-fitting cover in a desiccator. The crucible, etc., may be weighed as soon
as it has acquired the laboratory temperature (for a detailed discussion, see
Section in, 26 ).

XI, 14. CALCULATIONS OF GRAVIMETRIC ANALYSIS. The calcu-
lation of the weight of a constituent in a given precipitate follows directly, from
the proportion:

^n^p'-nMww.x

where M„Ap is the molecular weight of the precipitate, M the atomic (or

431

XI, 15/16 QUANTITATIVI: INORGANIC ANALYSIS

molecular) wciglu of the clement (or radical) sought, n the number of atomic (or
molecular) weights of M in the molecular weight ic is the weight of

precipitate, and ,v is the w eight of the conslitueni desired. Furthermore, if |(’is
the weight of the sample u.sed. the percentage of the constituent .sought;- is given
by:

.v; lK:;.v: KX)
or v — .V 100/ IF

Exiimplc. 1.000 gram of an iron compound, after .suitable treatment, yieMd
1). 1 565 g of ironflll) oxide. Calcul.ite the |xTceniage of iron in the compound,

Fc,0.,;2Fe .•:0.1565;.c

1 59,68 ; 2 K 55.84 ; : 0. 1 565 : .V
(Mol. wt. of Fc^O,) (2 .At. wt. of Fc)

V - ! -,lv’ !! d- 1 565 --- 0.6994 .< 0.1 565
b9.6S

-- 0. 1095 g of be
N'ow 0.1095; l,«i00:;y: 100
or y 10 95 /'er o//-e

Simple gravimetric determinations

XM5, DtrrFU.MlNArtO.N OF WATKR OF ttVDR.ATlO.N IN CRYS-
rALLKSFD UARIU.M CUI.ORIDH Barium chloride di-

hydrate loses all its waiter of crystallisation above KMl C. .Much higher
temperaiutc,s (up to S00-9(KI Clean be used m this dehydration, for anhydrous
barium chlortilc is mm-volalile and stable even at fairly iiigh temperatures,

BaCI;,2H/) - B.iCI,-i 211,0

With some hydrated s.ili,s. speci.il temperature limits must be observed.

Euni Jurc. lle.u a crucible and hd to dull redness for several minutes,
allow uj cool in a vlcsiccator. and weigh after 20 minutes. Introduce into the
crucible ! -1.5g of A.R, barium ehlorule. and weigh ag.iin. Place the covered
crucible, resting upon a pipc-cl.iy or silica tn.mglcCaboui [5cm above .a small
llame (nut more th.ui 5 -bem high). .At mtcivals of a few minutes increase live
liame gradually imiil the bottom of the crucible i.s heated to dull redness,
Mauuam theenieihle at tins temperature for about 1 U minme.s. allow it to cool in
a desiccator for 20 minutes, ami weigh. Rcpc.il the process until constant weight
(two consecutive weighings agreeing within 0.0002 g) is obt.iincd.

From the loss in weight, calculate the pereeniaee of water in barium chloride
diliydrate.

Similar determinations may be carricil out with magne.sium sulphate hep-
taphydrate (MgSOj.TH.O Mg.SO., i- 711,0). sodium tetraborate dccahydrate
(Na,B.,O7,!0H,O — iNa.B^O-A 1011,0), and vviih di-sodiuni hydrogen phos-
phate dodecahydrate(2Na,lH’Oj.l 211,0 ^ N’a^INOT -j-25H,0).

XI, 16. O rUER i)Kri:;RMIN.\TIO.N.S BY IGNI i lON. In addition to the
removal o( water ol hydration a number of oilier determinations may be carried

432

GRAVIMETRY XI, 17

out by simple ignition. These include iron in ammonium iron(III) sulphate
(‘ferric alum’),

(NH4)2S04.Fe2(S04)3,24H20 ^ Fe^Oa

aluminium in aluminium ammonium sulphate (‘ammonium alum’),

(NH 4 ) 2 S 04 . Al2(S04)3,24H20 ^ AI 2 O 3

bismuth in bismuth oxycarbonate (the residue obtained being the oxide Bi 203 ).
More detailed thermal analysis is carried out by the methods described in
Chapter XXIII.

XI, 17. DETERMINATION OF CHLORIDE AS SILVER CHLORIDE.
Discussion. The aqueous solution of the chloride is acidified with dilute nitric
acid in order to prevent the precipitation of other silver salts, such as the
phosphate and carbonate, which might form in neutral solution, and also to
produce a more readily filterable precipitate. A slight excess of silver nitrate
solution is added, whereupon silver chloride is precipitated :

Cl-+Ag+ =AgCl

The precipitate, which is initially colloidal, is coagulated into curds by heating
the solution and stirring the suspension vigorously; the supernatant liquid
becomes almost clear. The precipitate is collected in a filtering crucible, washed
with very dilute nitric acid, in order to prevent it from becoming colloidal
(Section XI, 8), dried at 1 30-150 °C, and finally weighed as AgCl.

Silver chloride has a solubility in water of 1.4mgdm“^ at 20 °C, and
21.7 mgdm"^ at 100 °C. The solubility is less in the presence of very dilute nitric
acid (up to 1 per cent), and is very much less in the presence of moderate
concentrations of silver nitrate (see Section II, 8 ; the optimum concentration of
silver nitrate is O.OSgdm”^, but the solubility is negligibly small up to about
1.7gdm“^). The solubility is increased by the presence of ammonium, and of
alkali metal salts, and by large concentrations of acids. Under the conditions of y
the precipitation, very little occlusion occurs. \J^f silver chloride is washed with^-'^
pure water, it may become colloidal and run through the filter) For this reason
the wash solution should contain an electrolyte (compare Sections XI, 3 and XI,

8). Nitric acid is generally employed because it is without action on' the
precipitate and is readily volatile; its concentration need not be greater thaiT^
O.OIAT. Completeness of washing of the precipitate is tested. for by determining
whether the excess of the precipitating agent, silver nitrate, has been removed.

This may be done by adding 1 or 2 drops of 0. 1 M-hydrochloric acid to 3-5 cm^ of
the washings collected after the washing process has been continued for some
time ; if the solution remains clear or exhibits only a very slight opalescence, all
the silver nitrate has been removed.

Silver chloride is light sensitive ; decomposition occurs into silver and chlorine,
the silver remains colloidally dispersed in the silver chloride and thereby imparts
a purple colour to it. The decomposition by light is only superficial, and is
negligible unless the precipitate is exposed to direct sunlight and is stirred
frequently. Hence the determination must be carried out in as subdued a light as

433

XI, 17 QUANTlTA'riVH INORGANIC ANALYSIS

possible, and when the solution containing the precipitate is set aside, it should be
placed in tlic dark (c.g,. in a locker), or the vessel containing it should becovered
with thick brown paper.

U has been ibiind that in a solution containing silver cliloride and 1-2 percent
e.vcess of 0.2.t/-sil'.er nitrate, exposure to direct sunlight for 5 hours with
occasional stirring leads to a positive error of about 2.1 percent wliilcc.xposurem
a briglit laboratory, with no direct or rcUccIcd sunlight and occasional stirring,
gives a positive error of about 0.2 per cent. This positive error is due to the
libcnition of chlorine during exposure to light: the chlorine is largely changed
back to chloride ions, which caif-e futihcr precipitation of silver chloride. A
possible re.iction is:

3CU-r5Ag' -i-allT) - 5.\ga C'lOj ' r fd 1'

On the other hand, in the deicrminaiion of .diver by precipitaliosi with a slight
e.xccss of(.),2.1Miidrochloric acid (Section .\I, 52), the error is negative, e.g.. 0.4
percent after 2 houtsexpo^uie in diicct sunlight with no stirring, and 0.1 jx'rccni
after 2 hours cxpo->ure in a bright laboratory, with no direct or retiecicd sunlight
and occasional stirring. This arises from the los.s of chlorine which escapes from
the precipitate. The weigiit of the precipitate tn.iy be brought to the correct value
by treatment wiili niine .icid. followed by hydrochloric acid.

Proc.\’dure. Weigh out accurately about U.2 g of the solid chloride (or an
amount containing appro.vimaieiy 0. 1 g of chlorine)* into a 250-'350cin^ beaker
pros idcd w itli .i stirring rex) and ciivercd w iih ;i clock glass. .Add about i 50em’ of
water, stir until the solid has dissolved, and ,u!d 0.5 env’ of concentrated nitric
acid. To the cold solution add 0.1. (/-silver nitrate slowly and with constant
stirring. Only a slight c.vcess sliould be added ; this i.s readily detected by allowing
the precipitate to settle and .uiding a few drisps of silv er nitrate solution, when no
further precipitate shoukl be obtained. Carry ou> the ih'Urttmulhm in subdued
/ig/ir. Heat the .susixmsion nearly to boUtng, while stirring constantly, and
maintain it at this temperature until the precipitate coagulates and the
supernatant liquid is clear (2 3 minutes). .Make certain that precipitation is
complete by additig a few drops of silver nitrate solution to the supernatant
liquid. If no further precipitate appeals, set the beaker aside in the dark, and
allow the .solution to .stand for iibout 1 hour before filtration. In the meantime
prepare . I filtering crucible (Goocli. porcelain or sintered ghi.ss -the last-named is
mc).si convenient); the crucible must be dried at litc same temperature as is
employed in heating the precipitate (13i)-150'C) and allowed to cool in a
desiccator (.see Section III, -(0, 41 fur details). Coileel the precipitate in the
weighed filtering ciueible (Section HI, 43). Wasli the precipitate two or three
tiinc-s by da'aniation with about 10cm-' v>f cold ver,‘ dilute nitric acid (say,
0.5 cm-' of the concentrated .icid added to ZOOeni^ of vvater) before transferriag
the precipitate to the crucible. Remove the la.si Muall particles of silver chloride
adhering to the beaker with a 'policeman' (Section 111, 27). Wa.sh the precipitate
in the crucible with very dilute nitric acid added in small portions (sec Sections
XI, S and HI, 42) until .5 -5cm^ of the washings, collected in a icsi-lube. give no

• A.R, poi;is',iuniors^),iiiinieh!()iHic.(Jric-J.a till 120 C. is -.uiuiblc.

434

GRAVIMETRY XI, 18

turbidity with 1 or 2 drops of O.lM-hydrochloric acid.* Place the crucible and
contents in an oven at 130-150 °C for 1 hour, allow to cool in a desiccator, and
weigh. Repeat the heating and cooling until constant weight is attained.

Calculate thepercentageofchlorine in the sample. ,

In this and all other gravimetric determinations, duplicate estimations are
recommended. Both determinations may be carried out simultaneously, or if this is
not convenient, the second should be commenced as soon as possible after the first is
in progress.

Note on the gravimetric standardisation of hydrochloric acid. The gravimet-
ric standardisation of hydrochloric acid by precipitation as silver chloride is a
convenient and accurate method, which has the additional advantage of being
independent of the purity of any primary standard (compare Section X, 37).
Measure out from a burette 30-40 cm^ of the, say, 0. 1 Af-hydrochloric acid which
is to be standardised. Dilute to 150cm^, precipitate (but omit the addition of
nitric acid), filter, and weigh the silver chloride. From the weight of the
precipitate, calculate the chloride concentration of the solution, and thence the
concentration of the hydrochloric acid.

XI, 18. DETERMINATION OF ALUMINIUM AS ALUMINIUM OXIDE.
Discussion. The aluminium is precipitated as the hydrated oxide by means of
ammonia solution in the presence of ammonium chloride. The gelatinous
precipitate is washed, converted into the oxide by ignition, and weighed as AljOj.

This determination is subject to several sources of error, most of which will now
be discussed. Aluminium hydroxide is amphoteric in character :

A1(0H)3-|-3H+ = A13+ -I- 3 H 2 O
or Al(0H)3-t-0H-+2H20 = [Al(0H)4(H20)2]-

Precipitation commences at approximately pH 4, and is complete when the pH
lies between 6.5 and 7.5. The latter pH range can be adjusted with the aid of
methyl red as indicator. The pH employed for precipitation must clearly be
controlled. This is achieved by the addition of ammonium chloride, which exerts
a buffering effect (Section II, 19) and also assists the coagulation of the initially
colloidal precipitate. The presence of ammonium salts reduces to a minimum
the co-precipitation of the divalent metals, such as calcium and magnesium (see
Section XI, 21) and other cations. A readily filterable precipitate is obtained by
precipitation in hot solution. The precipitate cannot be washed with hot water,
for the aluminium hydroxide is readily peptised (Section XI, 3), and will run
through the filter. A 2 per cent solution of either ammonium chloride or
ammonium nitrate is satisfactory; the presence of ammonium chloride in the
precipitate causes no appreciable volatilisation of aluminium during the
subsequent ignition (contrast iron(III) oxide).

A rapid method for drying the silver chloride, collected in a porcelain or sintered glass filtering
crucible, is as follows. (This method should not be used by elementary, students or beginners in the
study of quantitative analysis.) After washing the precipitate with very dilute nitric acid, wash the
walls of the crucible five or six times with small volumes of ethanol (a small pipette or a drawn-out
glass tube is useful for this purpose), and then several times with small volumes of anhydrous diethyl
ether. Suck the precipitate dry at the pump for 10 minutes, wipe the outside of the crucible with a
clean linen cloth, leave in a vacuum desiccator for 10 minutes', and weigh as AgCl. The procedure
may be employed for silver bromide, iodide, and thiocyanate. The results are usually slightly high.

435

XI, 19 OI^ANTITATIVB INOROAN'IC ANALYSIS

The aluminium oxide obtained by igniting aUiininiuin hydroxide is hy-
eroscopic unless the temperature has been raised to at least 1200 'C, when
apparently a non-hygroscopic form of the oxide i> formed. For tliis rea.sotithc
precipitate is ignited in a silica crucible (porcelain i.s slightly hygroscopic when
heated to a high temperature) over a Meker or Fisher burner, or with a blast
lamp. The best procedure is to linally heal for 10-15 nrinutes in an electric mufile
furnace at I2()0 C.

I’foccditre, Weigh out accurately about l.Sg of A.R. aluminium am-
monium sulphate (NH 4 ), .SO^.AljlSO^Jj, 2-111 ,U (or a weight of a sample
containing about 0.1 g of aluminium) into a 400- or OOO-cm^ beaker, provided
with a clock-glass cover and a stirring rod. Dissolve it in 200 cm^ of water, add 5g
of pure ammonium chloride, a few drops of methyl red indicator ( 0.2 per cent
alcoholic solution) ( 1 ). and heal just to boiling, .-ydd pure dilute ammonia
solution(! : Ijdropwise from a burette until ilic colour of the solution ch.mgs$ to
a distinct yellow. Boil the svdution for 1 or 2 minutes, and filter at once through a
suitable quantitative liiter paper (.Section III, .^ 8 i (2). Wash the precipitate
thoroughly with hot 2 per cent ammonium nitrate or chloride solution made
neulr.it with ammonia solution to methyl ted (or to phenol red). Place the paper
with the precipitate in a previously ignited silica or pl.itinuin crucible, dry, ch.ar,
and ignite lor 10 -1. S minutes with a .\leker or Fisher high-temperature burner.
Allow the crucible, covered with a wcU-futing lid, to cool in a dcsiamtor
containing a good desiccant, and weigh as soon as cold. Ignite to constant weight.

Calculate the (Krceiuagc ofaluminium in the sample.

Notes. 1. Phenol red {pH range: 6,4 (ycllowl to S.O (red)] has also been
recommended. 0.5cm^ of a 0.1 per cent solution of the indicator is added: the
colour cliange upon the addition of ammoni.i sohiiiuii is from yellow to orange.
Bromoctesol purple has also been used ; the purple end point ipH = 6 , 8 ) is taken.

2. .Asblc-ss paper pulp (.Section III, 39) may be added to assist the subsequent
filtration.

XI, 19. DOKK.MI.N.VriON OF Al.Li.MINTU.M AS ITIE S-HYDROXY-
QUI.NOi..VrE. Al(C,IIJ)Ni^. WlIll PKKCIPI l.Vl ION FRO.M IIOMO-
OENEOLS SOLUriO.N, Dimilwioh. Some of the dct.iils o! this
method have already been given in .Section XI, 1 1, C. This privcdurc separates
aluminium from beryllium (see. however, Section XI, 31, A), the alkaline earths,
magnesium, and phosphate. For the gravimetne determination a 2 per cent or 5
per cent solution ol o.vine m 2 .l/-.iceiic acid m.iy be used: Icm^ of the latter
Solution is suiliciciii to precipitate 3 mg of aluminium. For practice in this
determination, u.se .iboiii (1.40 g. accurately weighed, of A.R. aluminium
ammonium sulphate. Dissolve it in lOOcm'* of waicT, heat to 70 -SO C. add the
appropriate volume ol the o.vine reagent, and (if a precipitate h.is not already
lormed) slowly introduce 2 .lf-ammonium acetate solution until a prccipiluiejusl
appears, heal to boiling, and then add 25cnf* of 2,V/-amniuniinn acetate solution
dropwise and with constant stirring (to ensure complete precipitation). 11 the
supernatant liquid is yellow, enough o.vine reagent has been added. .Allow to
cool, and collect the precipitated aluminium ’oxinatc' on a weighed sintered glass
(porosity No. 4) or porous porcelain filtering crucible, and wash well with cold
water. Dry to constant weight at 1 .10- 140 Ck Weigh as A!(CglioON)j.

Precipitation may also be ctlected from homogeneoas solution. The c.vperimen-
lal conditions must be carefully controlled. The solution containing 25-50 mg of

436

GRAVIMETRY XI, 20

aluminium should also contain 1. 25-2.0 cm^ of concentrated hydrochloric acid
in a total volume of 1 50-200 cm^. After addition of excess of the oxine reagent,
5 g of urea is added for each 25 mg of aluminium present, and the solution is
heated to boiling. The beaker is covered with a clock glass and heated for 2-3
hours at 95 °C. Precipitationjs complete when the supernatant liquid, originally
greenish-yellow, acquires an orange-yellow colour. The cold solution is Mtered
through a sintered glass filtering crucible (porosity No. 3 or 4), washed well with
cold water, and dried to constant weight at 130°C.

Procedure. The solution should contain 25-50 mg of aluminium and
1.0-2.0cm^ of concentrated hydrochloric acid in a volume of 150-200 cm^. For
practice in this determination, weigh out accurately about 0.45 g of A.R.
aluminium ammonium sulphate, dissolve it in water containing about 1 .0 cm^ of
concentrated hydrochloric acid, and dilute to about 200 cm^. Add 5-6 cm^ of
oxine reagent (a 10 per cent solution in 20 per cent acetic acid) and 5 g of urea.
Cover the beaker with a clock glass and heat on an electric hot plate at 95 °C for
2.5 hours. Precipitation is complete when the supernatant liquid, originally
greenish-yellow, acquires a pale orange-yellow colour. The precipitate is
compact and filters easily. Allow to cool and collect the precipitate in a sintered
glass filtering crucible (porosity No. 3 or No. 4), wash with a little hot water and
finally with cold water. Dry at 130 °C. Weigh as A1(C9H60N)3.

XI, 20. DETERMINATION OF CALCIUM AS OXALATE. Discussion.
The calcium is precipitated as calcium oxalate CaC204,H20 by treating a hot
hydrochloric acid solution with ammonium oxalate, and slowly neutralising with
aqueous ammonia solution:

Ca^+ -1-02042- +H 2 O = CaC 204 ,H 20

The precipitate is washed with dilute ammonium oxalate solution and then
weighed in one of the following forms:

(i) as CaC204,H20 by drying at 100-105 °C for 1-2 hours. This method is not
recommended for accurate work, because, inter alia, of the hygroscopic nature of
the oxalate and the difficulty of removing the co-precipitated ammonium oxalate
at this low temperature. The results are usually 0.5-1 per cent high.

(ii) As CaCOa by heating at 475-525 °C in an electric muffle furnace. This is
the most satisfactory method, since calcium carbonate is non-hygroscopic.

CaC204 = CaC 03 -hC 0

(iii) As CaO by igniting at 1200 °C. This method is widely used, but the
resulting calcium oxide has a comparatively small molecular weight and is
hygroscopic; precautions must therefore be taken to prevent absorption of
moisture (and of carbon dioxide).

CaCOa = CaO + C02

Cakium oxalate monohydrate has a solubility of 0.0067 g and 0.0140 g dm” ^
at 25° and 95 °C respectively. The solubility is less in neutral solutions containing
moderate concentrations of ammonium oxalate owing to the common ion effect
(oection II, 9); hence a dilute solution of ammonium oxalate is employed as the
wash liquid in the gravimetric determination. Calcium oxalate being the salt of a

437

Xr, 20 QUANTITATIVE INORGANIC ANALYSIS

wc;ik acid, il-s solubility increases with iocreasinK hydrogen ion concentration of
Ihesoloiion because of the removal ofthc oxalate ions (compare Section 11, 12)to
form hydrogenoxalalc ions and oxalic acid:

CaC.Ojt (solid) Ca'* ’’ -f CjO** "

c\6r"fir^:5nc,04"

Calculation shosss that precipitation k quantitative at a pH of -1 or higher.

Precipitation from c»)Id neutral or ammoniacal solutions yields a veiy finely
div ided precipitate, vvhicit is diHicull to filter. Satisfactory results arc obtained by
adding amnionium oxahite to a hot m id solution ofihecaiciuin .s.ill (inorcorlcss
calcium oxalate may precipitate, deiveiuling upon the pH of the solution), and
finally neutralising with aqueou.s ammonia s»»lulion. The precipitate formed,
after digesting for about .m iuiuf, consists of relatively coar.se cry.stals which are
readily filtered. Better results are given by precipitation from homogeneous
solution according to the urea-hydrolysis method, details of which will be found
in Section X, 94 (sec also Section XI, 34).

In thk dctermin.iiion all those mct.ds (c.g.. copiser. lead, zinc) which form
sligiitly soluble c.x.datcs must be absent. Tiie problem which frequently arises in
practice is the precipitation of calcium in the presence of m.igne,siurn .ind the
alkali metals, The amount of the alkali metals which is precipitated i.s usually
.small; in the pre.seiice of large amounts t'f sodium, re-precipitation may be
desir.ible, Magiiesiutn may be eo-prccipit.ited (Section XI, 5) to a considerable
e.xicnt, but the amuuni of this m.iv I>e considerably reduced by not boiling the
solution and not allowing the ptecipitatc to stand in contact with the solution too
long before fiUraiion ipost-inccipitation. Section .XI, 5, is thus minimised). By
using a very large e.xeess ol'aiimionium oxalate, magnesium is held in solution in
the form of a complex s.di wuli the oxalate ion; furthermore, magnesium o.xalale
readily form,s quite stable supersaturated solutions. If the eoncenitalion ratio of
magnesium to calemm is e.xtremely large. ;i second precipitation is usually
necessary.

As already pointed out. calcunn. when preeipilatcd as oxalate, is best weighed
as the carbonate or o.xide. The theory of the decumpo.^ition of calcium oxalate is
ol sonic interest in tliis conticciion Decomposition of the oxalate into the
carbonate is rapid at about 42.s C. .At higher temperatuics. the disswiation of
calcium carbonate (CaCO^ ve:CaO-;-c6.) cvnucs into play. At any given
temperature, a mixture of CaCO^. C.tO.' and CO, in equilibrium with one
another exerts a certain delinitc pressure of carbon (.iioxide. If the partial pressure
ol the carbon dioxide in the surrounding atmosphere is gre.iter than the
equilibrium pressure for that tciiiperaiurc. the above reaction will prt'ceed from
right to left, and eventually tlic oxide will be comptclcly converted into the
carbonate. Ollierwisc expressed, calcium carbon.ite cannot be decomposed into
the oxide so long as the presNurc of carbon dio.xide in the surrounding
.itmosphere is greater than the equilibniiiu pressure ol the system
CaCO^-CaO-CO, at the temperature of heating. The atmosphere contains
iibout 0.f)3 per cent ol carbon dioxide by volume; when the pressure is 760 mm,
this corresponds to 760 s: (.).()003 0.22S[nm of mercury. Calcium carbonate

will tlieretore be perlcctly stable in the ainio.sphere so long as the decomposition
pres.siire doe.s not exceed 0.2.3 mm of mercury. The dissociation prcssure.s of

43.S

GRAVIMETRY XI, 20

calcium carbonate, expressed in mm of mercury, at various temperatures are
collected in the following table :

Thus calcium carbonate will not commence to dissociate appreciably in air until a
temperature of slightly above 500 °C is reached. Actual experiment has shown
that complete decomposition of calcium oxalate into the carbonate occurs at a
temperature between 475° and 525 °C; the rate of the decomposition
CaC 204 CaCOj + CO is slow at 450 °C but becomes reasonably rapid at
475 °C; above 530 °C the calcium carbonate commences to lose carbon dioxide.
For the weighing of calcium oxalate as calcium carbonate, fine temperature
control is necessary ; this can be achieved only by the use of an electrically heated

muffle furnace, provided with a pyrometer or suitable thermometer. If such
equipment is available, the method should be used in preference to all others ; the
oxalate must be filtered through a Gooch (preferably of silica) or porcelain
filtering crucible and not through filter paper.

Above 882 °C calcium carbonate is completely decomposed into the oxide, but
unless the carbon dioxide is removed by diffusion, convection, etc;, by
conducting the ignition in a loosely covered crucible, there will be a re-
combination of calcium oxide and carbon dioxide on cooling, with the formation
of some calcium carbonate. In practice, it is found that the rate of decomposition
at about 900 °C is very slow, and it is best to use a temperature of 1100-1200 °C.
This temperature is not easily attained in a porcelain or silica crucible unless an
electrically heated muffle furnace is employed. However, quantities up to 1 g can
be completely decomposed in a platinum crucible by the use of a Meker or Fisher
high-temperature burner. The residue of calcium oxide is hygroscopic (unless
heated for a considerable time above 1200 °C). The crucible should be kept well-
covered in a desiccator, containing pure concentrated sulphuric acid, freshly
Ignited quicklime, or phosphorus pentoxide, and weighed immediately it has
acquired the laboratory temperature. Although anhydrous calcium oxalate

appears to be stable between 226 °C and 398 °C, this is not used as a weighing
form because of its hygroscopicity.

Procedure. Weigh out accurately sufficient of the sample to contain 0.2 g
of calcium* into a 400- or 600-cm^ beaker covered with a clock glass and
provided with a stirring rod. Add 1 0 cm^ of water, followed by about 1 5 cm^ of
dilute hydrochloric acid (1 : 1). Heat the mixture until the solid has dissolved, and
boil gently for several minutes in order to expel carbon dioxide. Rinse down the
sides of the beaker and the clock glass, and dilute to 200 cm^ : add 2 drops of
methyl red indicator. Heat the solution to boiling, and add very slowly a warm
solution of 2.0 g of ammonium oxalate in 50 cm^ of water. Add to the resultant
hot solution (about 80 °C) filtered dilute ammonia solution (1 1 1) dropwise and

0.5 gram of A.R. calcium carbonate, or of calcite, which has been finely powdered in an agate
mortar and dried at 110-130°Cfor I hour, is suitable.

439

XI, 21 QUANTITATIVE INORGANIC ANALYSIS

with stirring until the mixture is neutral or faintly alkaline (colour change from
red to yellow). Allow the .solution to .stand without further heating for at least an
hour. After the precipitate ha.s settled, lest the .solution lorconiplete precipitation
with a few drops ofainnionium oxalate solution. The subsequent procedure will
depend on whether the ctilcium oxalate is to be weighed as thccarbonatcorasihe
oxide.

Weighing as calcium earhonale. Decant the clear .supernatant liquid
through a weighed silica Gooch crucible ora porcelain filtering crucible. Transfer
the precipitate to the crucible with a jet of water from the wash bottle; any
precipitate adhering to ilie beaker or to the stirring rod is transferred with the aid
of a rubber-tipped rod (‘policeman’). Wash ib.e precipitate with a cold, very
dilute ammonium oxalate solution (0.1 -0,2 per cent) at least five limes, or until
the washings gi\c no Ic-si for chloride ion (add dilute nitric acid and a few drops of
.silver nitrate solution to 5 cm-* of the washings). Dry the precipitate in tlie steam
oven or at iUO- !2I) C fi.>r 1 hour, and tlicn transfer to an electrically heated
mul]Ic I'urnace, maini.uiied at 500x25 C lor 2 iioufs. Cool ilte crucible and
contents in a de.sicealor.aiid weigh. I-urihcr hc.umg at 500 Cshould not alTeet the
weight. As a linal precaution, moisten the prceipilaie with a few drops of
saturated ammonium carboiiale solution, dry at 1 10 C, and weigh again. Again
in weight indieatc.s that sitmc oxide was ptcsctii: this should not wcur.

CaO lTNTUl.CO, - CaCO., f2NH,Tll.(>

Calciiiaie the iwrceniagc of calcium m the sample.

Weighinj* us calcium oxide. Decant the cle.ir. .siipernataiu liquid through a
Whatman No, •!() or 5-10 filler paper, iran.sfcr the precipitate to the filter tSeclion
111,38), and wash with a cold 0.1 -0.2 percent .immonium oxalate .solution until
free from cliloride. Tran.sfcr the moist precipitate to .i jircviously ignited and
weighed platinum crucible, and ignite gently .it over a Bunsen ilamc ami
finally for 10 -15 mimites with a .Mcker or b'lshcr iiigU-tcmpcrature burner until
two successive weighings do not dili'er by more ih.m 0.0003 g. The covered
crucible and contents are piaccri m a desiccator containing pure concentrated
sulphuric acid vw phosphorus pcntiixide (but not calcium ciiloride), and weighed
as soon as cold.

Calculate the pcrccniiigc of ealcium in the s;imp!c.

For Ollier meihod.s for the detcrmin.iiion of calcium, including precipitation
from homogcncou.s .solution, sec Section ,X1, 34.

.\I, 21. DETERMI.N.VTIOiN OF IRO.N AS IRONllH) OXI Dlv DiiOi^sioii.
The solution containing the iron(iU) s.ilt* i.s treated with a slight
excess of aqueous ammonia solution to precipitate the hydrated oxide
f-e,Oj,.vH,0. Ihe precipitate has no definite stoichiometric composition, but
contains a variable amount of water, partly bound chemically and partly-
adsorbed, It is convenient, however, in writing ctjuaiions for reactions involving
hydrated oxides and also for c.ileulatmg solubility product constants, to assume
the hydroxide lormula. although, m most cases, the composition ol the
precipitate does tun correspond to this formula. The equation for the

• IroiidDusonl) parii.iiU prccipiutcd hv .imiiumu solutiou in ihc pwsence of ammonium i-alU-

GRAVIMETRY XI, 21

precipitation of hydrated iron(III) oxide may be written as:

[Fe(H20)6]^+ + 3 NH 3 = Fe(H20)3(0H)3 + 3 NH 4 +
or as , Fe^+ +3NH3 + 3H2O = Fe(OH)3 + 3NH4+

Other elements that are precipitated by ammonia solution must of course, be
absent. These include aluminium, trivalent chromium, titanium, and zirconium.
In the presence of an oxidising agent (even atmospheric oxygen) manganese may
be precipitated as the hydrated dioxide. Anions, such as arsenate, phosphate,
vanadate and silicate, which yield insoluble compounds of iron in weakly basic
media, must be absent. The presence of salts of organic hydroxy-acids (e.g.,
citric, tartaric, and salicylic acids), hydroxy compounds (e.g., glycerol and
sugars), alkali pyrophosphate and fluorides must be guarded against, because of
the formation of complex salts and the consequent non-precipitation of iron(III)
hydroxide.

The solubility product of iron(III) hydroxide is of the order of 10 ~^®, so that
quantitative precipitation occurs even in weakly acid solution, and errors due to
washing will be negligibly small. The precipitate first forms as a dispersed phase,
but on heating in the presence of electrolytes it coagulates to a gelatinous mass,
which settles out of suspension; prolonged heating tends to break up the
aggregates and causes the precipitate to become slimy. The hydrated iron(III)
oxide is a typical example of a flocculated colloid. The coagulation of a colloidal
precipitate, and especially the agglomeration of the primary particles, is aided
considerably by raising the temperature of the solution. Hence precipitation is
carried out at or near the boiling point, and the liquid is maintained at this
temperature for a short time after precipitation.

As might be expected from its colloidal character, hydrated iron(III) oxide has
a great tendency to adsorb other ions present. If precipitation is made from basic
solution, the primary adsorbed ion is the hydroxide ion (Section XI, 3 ), and this
readily holds by secondary adsorption positive ions which may be present. If
there is a large excess of ammonium ions in the precipitating and wash solutions,
the adsorption of other cations can be kept at a minimum ; since ammonium salts
are volatilised upon ignition of the precipitate, little harm is caused by their
adsorption. Divalent ions are more strongly adsorbed than monovalent ions
(Section XI, 5 ). If the extent of co-precipitation is large, purification may be
effected by re-precipitation, since the precipitate is soluble in dilute acids.

The gelatinous precipitate of iron(III) hydroxide is always filtered through
filter paper. Application of suction, in order to hasten filtration, should not be
attempted, since the effect of the suction is merely to force the small particles of
the precipitate into the pores of the filtering medium. It may often happen that
with suetion the liquid will pass through more rapidly ; this does not mean that
the washing process is accelerated, since the liquid runs through small channels
and does not permeate the main body of the precipitate. For this reason iron(III)
hydroxide is best washed by decantation ; the precipitate may then be thoroughly
stirred with the wash liquid. To prevent peptisation and the production of slimy
luaterial, an electrolyte is used in the wash liquid. The most satisfactory is
ammonium nitrate; this volatilises upon ignition and assists somewhat in the
subsequent ignition of the precipitate. Ammonium chloride is unsuitable,
because iron(III) chloride, which is volatile, is formed during the ignition ;

Fe303 + 6NH4C1 = 2FeCl3-t-6NH3-I-3H20

441

XI, 21 QUANTITATIVE INORGANIC ANALYSIS

It is advisable, therefore, to wash out nearly all tlie animonium chloride present
in the hydrated iron(l II) oxide; very .small amounts, however, will not lead to any
sisinificanl error. To assist filtration, a hot wa.sh solution should be employed.
The filtration and washing of any gelatinoii.s precipitate is hastened by the use of
;ishle.ss liiter pulp (Section 111, 39). Under no circumstances should the
precipitate be allowed to stand in the filter paper before washing is complete,
because it shrinks rapidly as it partially dries, and channels, which permit the
wash liquid to run through, are formed in the precipitate.

Hydrated ironllll) oxide upon ignition at lOOtJ C yields iron(lll) oxide; at
higher tempcr.iturcs tri-iron letroxide i.s slowly formed. The ignition should be
carried out under good oxidising conditions, especially during the burning of the
filter paper, for otherwise partial rcduelion to the magnetic oxide FciO.;, or even
to the metal, may occur. These reduction products arc only slowly converted into
iron(!ll) oxide upon continued heating with ftce access of air, Such reduction is
avoided by burning oif the ciuboii at a low heat, by maintaining at all times free
access of air, and by excluding the reducing gmes from the llame.

Prmai'itri.’. For pr.ictice in this dclerminaiion, the .student may employ
cither A.R, ammonium iron(ll) .sulphate (.NH.j);S04 FeS04,6{l,0 or A.R.
ammonium iron(lU) sulphate (Nll4)..S04Fc,{S04)j,2-}H.0. The former is to
be preferred, as this dctermin.ition insolvcs oxidation of the iron(ll) salt to the
iron(Ull stale. Weigh out accurately about 0..S g of ammonium iroitdl) sulphate
into a -KHJ-cnF beaker ptov iilcd with a clock glass and stirting rod. Dissolve it in
SOcnF of water and lOcnF of dilute hydrochloric acid (111), Add l--2cnF of
concentrated nitric acid* to the .solution, and boil gently until the colour is clear
yellow (.T-5 minutes is usually uccc''sary) (1). Dilute the .solution to 200cra^i■
heat to boiling, and slowly add pure I : I ammonia solution (2) in a slow stream
from a small beaker until a slight excess is present, a.s is shown by the odour of the
steam above the liquid (3). Hoil the liquid gently for I minute, and allow the
precipitate to settle. The supernatant liquid should be colourless. As soon us
most of the precipitate has sctilevl, decant the sui>crualaiU liquid through an
ashiess filler paper, hut leave as much of the precipitate as possible in the beaker.
It is essential that the liiter p.ipcr tits tiie funnel properly (Section III, 3S), so that
the stem of liie liinncl is always filled with liquid, otherwise liltration will be very
.slow. Add about lOUcnF iif liuiling 1 per cent ammonium nitrate solution to the
precipitate, stir the mixture thoroughly, and allow to .settle. Decant as much
liquid as possible through the filter. Wa.sh the precipitate three to four times by

* Thcrcjclion a:

3Fc'* , NO,’ r ill' .. 3Fe*" + NO r’ll.O

Ths ai'..ni'..iniai;c o!" this pnxtdisic i, Ui.ii ilu- proeiice ,>1 nilralci is undent, iWc if sulph.ilc ir
suF%ct|UciuK lo he delennined ni Ihc liUr.iie as IhrSO,, ncecvsilalnit* one or more cv,iptuJlions lo
drsnc's vuih !i>dtoch!v>rie .leul lo remove the tiiiiic .icnl this ilillieiiily m.iy he avoided by
etiiployniit ciiher br.imtnc w.iler (dFc'' rUr. ..;Fc** rdllr'i or hvdioscn perodde for l!ic
oodulion. Add 10- l5t!n’ofs,(tur.ued btinumevvalfrusi!seiuiisohiuonii-c.,nvniovier.iicevce:^JS
indic.iicd by lie eolour of ibe soUilioii ,ind (he pervisteivl odour v't bsotmiie - e,iuUon!) and boil lo
coinplcle tlie ovidalion .nid !o remove tiiovl of the eveevs of bromine. Iljdropen perovide is
eoii\emcnllycnip!o)ed ,is ihe •lOO-volumc solunon. Add t e!trioftbc!.(iler.inddcsirovlhec\ee>s
of re,igcni by boiliiH!.

t If A.R (ammonium irondll) sulph.iicl is used, dissobc h.'l g (or a sullicicnl amouni of an
iion(lll) v.ili eonlaiiimg about 0. l.s y of iron) m dOOcuri of u.ncr, add lOeiid of ! : 1 iijdnwhloric
aeid. and pnxeed .is described.

GRAVIMETRY XI, 21

decantation with 75-100-cm^ portions of hot 1 per cent ammonium nitrate
solution. Transfer the precipitate (and ashless filter pulp, if employed) to the filter
(Section HI, 39) ; any small particles adhering to the sides of the vessel or to the
glass rod are dislodged with the aid of a ‘policeman’, and subsequently
transferred to the main precipitate with the assistance of hot water from a wash
bottle. Wash the iron(III) hydroxide several times with hot ammonium nitrate
solution (4) until no test (or, at most, a very slight test) for chloride is obtained
from the washings. Allow each portion of the wash liquid to run through before
adding the next portion; do not fill the filter more than three-quarters full of the
precipitate. While the filtration is in progress, ignite a clean crucible (porcelain,
silica, or platinum) at a red heat, cool in a desiccator for 20 minutes, and weigh.
When the filter paper has drained thoroughly, fold over the edges, and transfer to
the weighed crucible. Proceed as described in Section XI, 18. Heat gradually until
dry, char the paper without inflaming, and burn off the carbon at as low a
temperature as possible under good oxidising conditions, i.e., with free access of
air in order to avoid reduction of the iron(III) oxide. Finally, ignite the
precipitate at a red heat for 15 minutes and take care to exclude the flame gases
from the interior of the crucible, cool in a desiccator for 15 minutes, and weigh.
Alternatively, heat in an electric muffle furnace at 500-550 °C. Repeat the
ignition (10-15 minues) until constant weight is obtained (to within 0.0002 g).

From the weight of iron(III) oxide obtained, calculate the percentage of iron in
the salt used.

Notes. 1. At this stage test the solution for the complete oxidation of the
iron. Transfer a drop of the solution to a test tube by means of a stirring rod, and
dilute with about I cm^ of water. Add a few drops of a freshly prepared
potassium hexacyanoferrate(III) solution (0.1%). If a blue colour appears
iron(II) is still present in the solution, and more nitric acid must be added.
Alternatively, use a drop or two of 0.1 per cent aqueous 1,10-phenanthroline:
iron(II) gives a red colour.

2. Filtered ammonia solution should be used in order to prevent the
introduction of silica, which is often present in suspension in alkaline solutions.

3. At this point it is advantageous to add a little ashless filter pulp, best in the
form of a Whatman ‘accelerator’ or one-fourth of an ‘ashless tablet’. For further
details, see Section HI, 39.

4. If desired, hot water from a wash bottle may be substituted at this stage ;
peptisationis negligible.

It is interesting to note that if a few drops of hydrazine hydrate are added
immediately after the ammonia solution and the suspension boiled for 30-60
seconds, the precipitated iron(III) hydroxide is in a relatively compact form and
filters fairly easily. The precipitate may be filtered through a Whatman No. 541
filter paper, washed with 1 per cent aqueous ammonium nitrate solution, and
finally three times with warm water. After charring the filter paper as described
above, the precipitate is heated at 450 ‘’C, cooled, and weighed as FejOj. This is
an improvement on the conventional method of precipitation using ammonia
solution alone. The reader may wish to repeat the determination using this
modified procedure and compare the results obtained.

Iron(III) can be precipitated from homogeneous solution as a dense basic for-
mate by the urea hydrolysis method (compare Section XI, 7). The basic iron(III)
ormate is easily filtered and readily washed, and adsorbs fewer impurities than
he precipitate obtained by the ammonia and other methods. Ignition of the basic

443

XI, 22/23 QUANTITATIVE INORGANIC ANALYSIS

formate yields iron(lll) o,\idc. For experimental details of this and other
methods, see Section XJ, 40.

XI, 22. determinat ion OF LE.-VD AS CHROMATE, Dixeussion.
Although this method is limited in its applicability because of the general
in.solubiiity of chromates it is a useful procedure for gaining experience in
eravimctric analysis. The best results arc obtained by precipitation from
homogeneous solution utilising the homogeneous generation of chromate ion
produced by slow oxidation of chromiumdll) by bromatc at 90-95 C in the
presence of an acetate bulfer. For further details see Section XI, 36B.

Froixthirc, Use a sample solution containing0.l ~0.2g lead. Neutralise the
solution by adding sodium hydro.xide until a precipitate just begins to form. Add
10cm' acetate bulfer solution (6,1/ in acetic acid and 0.6,1/ in sodium acetate).
Idem' chromium nitrate solution (2.4 g per 100cm'), and 10cm' potassium
bromatc solution (2.0 g jver lOOcm'l. Heat to 90-95 X'. After generation (of
chromate) and precipitation are complete (about 45 minutes) a.s shown by a clear
.supernatant liquid, cool, filter through a weighed sintered glass or porcelain
lilieringcrueible.wash with a little 1 [K-rcent nitric acid, ami dry at 120 C.Wci2h
as PbCrO...

XI. 23. DIvIEK-MINAlTON OF .MAGNESIUM AS THE AMMONIUM
I’HOSiTIATE HEXAHVDR.VTE AND AS THE PYROPHOSPH.ATE.
Oixaixxwn. A cold acid solution rif the magnc-siuin salt i.s treated with an exce.ss nf
diammonium hydrogenpho.spliaic. and then excess of ammonia solution is
added to (uecipiiate ammonium magnesium phosphate he.xahydrale,
MgNiUFO^.f'HjO, at room temperatuiet*

Mg'' •f-Hl’04'- s-NTU’ rOH -- .MgNH^PO^-rH.O

This precipitate po.sscssesa relatisely iiigh solubility I about 65 mgdm“'at 10 'C
in pure water, but less m the presence of dilute .tqueous ammonia), and it also has
a tendency to form supersaturated solutions; the svsluiion should therefore stand
for several hours before filtration. T he precipitate is washed with O.S,l/aqucou,s
ammonia solution^ (say, i.'191 and then weighed cither as the he.xahydrale
.MgNi UPOj.blUQ or as the pyrophosphate ,Mg,P>0..

For the former, the precipitate is washed with ethanol, followed by anhy drous
diethyl ether, and weighed after standing at room temperature (preferably in a
desiccator) for about 20 minutes. Thi.s method i.s of modenitc accumey. and is
recommended here bec.iuse of dillicuities which attend the ignition of the
precipitate and the time saving achieved.

•The prccipil.itioii slioutit be carncJ out at IS ,ni C m etCer la cusuic the .ibscncc of the
nioni)h\tlraic, i .0. 7 tie Uucrs.ilt loniw.inO is si.jhEc iu s.’iulii'us abiuc 62' C; uhen

onec fofnv. J, it ukes jboui 21 tuuir. sCinOui^ at tuom icaiperaiiirc bcfoie ii is eonverteO into the
tic.x.ilijdr.ilc

r 7 he appiovnn.ilc u,lubilil:o. ctprcs'cd js nn; of.MtjNl QI'Oj.ull .Otlm ' al room Icmpewlure in
aqueous .immonu soluiunw of t.itious coiiccnttaiious .iic 6l2.\f, !2; 0..’.tf, 6; 0.6,tf, 3:
1.2.\/. 1; 1.7A/,0S.

GRAVIMETRY XI, 23

For the latter, the precipitate is ignited at a high temperature (>1000 °C for
1 hour) to magnesium pyrophosphate and weighed as such:

2 MgNH 4 P 04 = Mg 2 P 207 + 2NH3 + H20

To obtain a precipitate of the correct composition (MgNH4P04) at the first
precipitation is a difficult matter owing to the co-precipitation of ammonium
phosphate and magnesium, pimsphates ; however, if the experimental conditions-
are carefully chosen and a pure magnesium salt is used, a precipitate of normal
composition is formed. If the precipitation takes place in the presence of much
ammonium salts, the precipitate may contain Mg(NH4)4(P04)2 ; the latter gives
magnesium metaphosphate Mg(P03)2 upon ignition. If the precipitation is made
in the presence of much potassium or sodium salts, the precipitate is con-
taminated with magnesium potassium (or sodium) phosphate. Hence if much
ammonium, potassium, or sodium salts are present, re-precipitation is essential.
In any case, re-precipitation is desirable to procure the best results. The double-
precipitation process will accordingly be described. It is an experimental fact that
the precipitate is practically insoluble in 5 per cent ammonia solution; this is
accordingly used as the wash liquid.

Great care must be taken in the conversion of ammonium magnesium
phosphate into the pyrophosphate. The carbon must be burnt off" at as low a
temperature as possible, because of the danger of the reduction of the phosphate
precipitate if the heating is strong while carbon remains ; if a platinum crucible is
used, the resulting phosphorus may lead to serious damage to the crucible.
Furthermore, if the heating is rapid, a dark-coloured product is obtained. For
these reasons, the charring of the paper and the burning off of the carbon are
conducted at as low a temperature as possible ; the temperature must be raised
very gradually. Some authors recommend, particularly for elementary students,
that the filter paper be ignited apart from the precipitate (Section III, 44) in order
to minimise this danger. It is preferable, however, to collect the precipitate in a
porcelain filtering crucible; this is then heated in an electric muffle furnace at
1000-1 100 °C.

Procedure. To a neutral or slightly acid (hydrochloric) solution of a
magnesium compound, containing not more than 0.1 g of magnesium,* add
5cm^ concentrated hydrochloric acid, and dilute to 150cm^. Add a few drops of
methyl red indicator to the cold solution, and then lOcm^ of the freshly prepared
ammonium phosphate reagent (25 g of A.R. (NH4)2HP04 dissolved in lOOcm^
of water). Now add pure concentrated ammonia solution slowly while stirring
the solution vigorously until the indicator turns yellow. Avoid scratching the
sides of the beaker with the stirring rod, for wherever there is contact, an
adhering crystalline deposit forms quickly. Continue to stir the solution for 5
minutes, adding ammonia solution dropwise to keep the solution yellow, and
finally add 5 cm^ concentrated ammonia solution in excess. Allow the solution to
stand in a cool place for at least 4 hours or preferably overnight. The precipitate
may be weighed either as MgNH4P04,6H20 or as Mg2P207.

About 0.6g of A.R. magnesium sulphate, accurately weighed, is a convenient quantity of
niagnesium salt to employ for this estimation.

445

XJ, 23 OUANTITATIVK INORGANIC ANALYSIS

Weighing as MgNH 4 p 04 . 6 Hj 0 . Filler through a sinicrcd glass or
porcelain filiering crucible which has been wushed with ethanol and diethyl ether
and weighed. Wash the precipitate with small portion.s of dilute ammonia
solution (1:19; ai. O.SAf) until a few ciif* of the filtrate, when acidified with dilute
nitric acid and tested with silver nitrate solution, give.s no test for chloride. Now
wa,sh with three lO-cm^ portions rectified spirit (95 per cent ethanol), draining
well after each washing; thi.s serve.s to remove most of the adhering water. Finally,
wash with five 5cm^ portions of anhydrous ilieihyl ether, draining after each
wa.ihing. Then draw air through the crucible for 10 minutes, wifse the outside of
the cold crucible with a clean linen cloth, and allow to stand in the air or in a
de.siccator for 20 minute-S. Weigh as .MgNH^PO^.filFO.

Weighing as Mgil’iO,. Filter ihroiigli a porcelain filtering crucible,
taking great care to remote all the precipitate from the beaker and stirring rod.
wash with .small portions of cold 0.!<3/-aqucous ammonia solution until the
washings give no turbidity witli dilute nitric acid and silter nitrate solution. Dry
the filtering crucible in an air o\en .it 100 150 C for an hour, and then heat it
gradually in an electric mulllc furnace to 1000 ■ 1 100 C. and .maintain it at this
temperature until constant weight is attained. If an electric furnace is not
available place the porcelain filter crucible inside a nickel crucible (or u.sc the
Ignition dish supplied with the crucible), and then beat gradually to the full heat of
a Meker, Fisher, or equisalenl burner. Heat for 25-30-minule periods until
constant weiglif is attained. Weigh as .Mg.P.O,.

AliernaOsely, but less .satisfactorily, the precipitate may be filtered through;!
ciuaniilaiivc filter paper. Wash tlie precipitate on the paper with cold O.S.U-
aqueous ammonia solution until the washings gi\e no turbidity with dilute nitric
acid and silver nitrate solution. Dry the precipitate at 100 C and place it in a
previou-sly ignited and sveighed platinum crucible (I), Char the paper slowly
without allovsiiig it to ignite, and burn oif the carbon .it as low a temperature as
possible With free access of air (gradually incfea.se the flame, but do not heat the
crucible to more than the faintest red), and then ignite to constant weight in an
electric imillle furnace.it 1000-1 100 Cor, less desirably, over a Meker or Fisher
high-temperature burner.

Caieuiaic the percentage of m.ignesium m the cuinpound.

If there is tune for a second prectpiiaiion. or if a pure magnesium salt is not
used and consequently the purity of the precipitate may lx* suspect, it is adv isabic
to carry out a .second precipitation. In thi.s ca.se the initiid pra'ipiiaie is
conveniently collected on a iiuantilativc filter paper (Whatman No. -10 or No.
540): a little 0.1S,t/-aqueou.s ammonia solution should he u.scd to assist the
transfer of most of the prccipii.ile to the filter paper. Dis,solve the precipitate on
the filter paper in appro.simately 50 cim* of warm dilute h> drochioric acid (DIO),
and wash the paper thoroughly with hot very dilute hydrochloric acid (DlOO)
into the beaker used for the initial precipitation. Dilute to 125 -1 50 cnC. add a
few drops of methyl red indicator, (J..3g of .A.K. di.immonium hydrogen
phosphate, and again precipitate the ammonium magnesium phosphate by the
addition of concentrated ammonia solution dropwise (preferably from a burette)
and vviiii constant stirring until the solution is yellow, followed by a further 5cm^
of the ammonia solution. Allow the solution to stand for at least 4 hours or,
better, overnight. Weigh the precipitate a.s MgNU or as MgiPA as

detailed above. Magnesium may also be determined as the ,S-hydro.vyquinal-
dinate; .see Section ,\I, 43.

446

GRAVIMETRY XI, 24

Note. 1. For elementary students it is sufficient to dry the filter with the
precipitate in the steam-oven (or at 100“C), and to incinerate the filter paper
apart from the precipitate (Section III, 44) at as low a temperature as possible;
the paper should not be allowed to take fire. After the volatile carbonaceous
matter has been burnt off, the residue may be ignited strongly with the lid of the
crucible displaced, to allow, circulation of the air, until the residue is as white as
possible. The main precipitate is added, and the whole ignited to constant weight
as described above.

XI, 24. DETERMINATION OF NICKEL AS THE DIMETHYL-
GLYOXIMATE. Discussion. The nickel is precipitated by the addition of
an ethanolic solution of dimethylglyoxime {CH 3 -C(:N 0 H)-C(:N 0 H)-CH 3 ,
referred to in what follows as H 2 DMG} to a hot, faintly acid solution of the
nickel salt, and then adding a slight excess of aqueous ammonia solution (free
from carbonate). The precipitate is washed with cold water and then weighed as
nickel dimethylglyoximate after drying at 1 10-120 °C. With large precipitates, or
in work of high accuracy, a temperature of 150°C should be used: any reagent
that may have been carried down by the precipitate is volatilised.

NP++ 2 H 2 DMG = Ni(HDMG)2 + 2H+

(For the structure of the complex and further details about the reagent, see
Section XI, llA).

The precipitate is soluble in free mineral acids (even as little as is liberated by
reaction in neutral solution), in solutions containing more than 50 per cent of
ethanol by volume, in hot water (0.6 mg 100 cm "^), and in concentrated
ammoniacal solutions of cobalt salts, but is insoluble in dilute ammonia solution,
in solutions of ammonium salts, and in dilute acetic acid-sodium acetate
solutions. Large amounts of aqueous ammonia and of cobalt, zinc, or copper
retard the precipitation; extra reagent must be added, for these elements
consume dimethylglyoxime to form various soluble compounds. Better results
are obtained in the presence of cobalt, manganese, or zinc by adding sodium or
ammonium acetate to precipitate the complex; iron(IIIX aluminium, and
chromium(III) must, however, be absent.

Dimethylglyoxime forms sparingly soluble compounds with palladium,
platinum, and bismuth. Palladium and gold are partially precipitated in weakly
ammoniacal solution; in weakly acid solution palladium is quantitatively
precipitated and gold partially. Bismuth is precipitated in strongly basic solution.
These elements, and indeed all the elements of the hydrogen sulphide group,
should be absent. Iron(II) yields a red-coloured soluble complex in ammoniacal
solution and leads to high results if much of it is present. Silicon and tungsten
interfere only when present in amounts of more than a few milligrams. Iron(III),
aluminium, and chromium(III) are rendered inactive by the addition of a soluble
tartrate or citrate, with which these elements form complex ions.

Dimethylglyoxime is almost insoluble in water, and is added in the form of a 1
per cent solution in rectified spirit or absolute ethanol; 1 cm^ of this solution is
sufficient for the precipitation of 0.0025 g of nickel. As already pointed out, the
reagent is added to a hot feebly acid solution of a nickel salt, and the solution is
then rendered faintly ammoniacal. This procedure gives a more easily filterable
precipitate than does direct precipitation from cold or from ammoniacal
solutions. Only a slight excess of the reagent should be used, since dimethyl-

447

XI, 24 QUANTITATI Vli INORGANIC ANALYSIS

glyoxime is not very .soluble in water or in very dilute ethanol and may-
precipitate; il'u \ery huge e.xcess is added (such that the alcohol content of the
solution exceeds 50 percent), some of the precipitate may dissolve.

Prua’durc. S- Nickelin a nickel .salt. Weigh out accurately 0,3-0.4g of
pure (preferably A.R.*) amnioniuin nickel sulphate (NH4)jSO4-NiSO4,6Hj0
into a •100-cni-' beaker provided with a clock-glas.s cover and stirring rod.
DLssolve it in water, add 5cnv* of dilute hydrochloric acid (III) and dilute to
200cni-’. Heat to 70-S0 C, add a slight c.xccss of the dimethylglyoxinie reagent
(at least 5cin-’ for every 10 mg of Ni present), and immediately add dilute
ammonia solution dropwise, directly to the .solution and not down the beaker
wall, and with constant stirring until precipitation takes place, and then in slight
C.XCC.SS. .Allow to stand on the steam bath for 20- 30 minutes, and test the solution
for complete precipitation when the red precipitate lias settled out. Allow the
precipitate to .stand for 1 hour, cooling at the -.amc time. Filter the cold solution
through a (iooch. sintered gl.iss or porcelain filtering crucible, previously heated
10 110-120 C and weiglied after cooling in a desiccator. Wash the precipitate
willi cold water until free from chloride, and dry it at 110-120 ‘C for 45-50
minutes. Allow to cool in a desiccator and weigh. Repeat the drying until
con-stunt weight is attained. Weigh as Ni(Cil UO,N.),. which comaims 20.32 per
cent of Ni.

Calculate the percentage of inckcl in the .salt.

B. Nickel in nickel steel. Weigh out accurately about I g of the drillings or
bonng.s of the nickel .steel t (or sutlicieiu of the sample to contain 0.03- 0.04g of
nickel) into a 101) - ISO-cm’ beaker or poicelain b.tsin, dissolve if iniheminiimim
volume of concent rated hydrochloric acid (about 20cnH slioukl suliice). and boil
with succe-ssive additions of concentrated nitric acid (nr. Scin^) to ensure
complete oxidation of tb.e iron to the iron(lU) state. Dilute somewhat, filler, if
necessary, from any solid material, and wash the paper with hot water: dilute the
filtrate (or soluliun) to 250 enH in a 4tj(j.em-* beaker. .Add 5 g of citric or tartaric
acid, neutralise the solution wuh dilute .uiu-eous ammonia solution, j and then
barely acidify (litmus) wuh dilute hydrochlonc acid. Warm the solution to
60 si) C, add a slight excess of .i 1 per cent clhaiiolic solution of dimethyl*
glyoxmie (20 25ciiH). iimueduilely rollowcd by dilute ammonia sokilion
dropwise until the liquid is slightly amniom.ical. stir well, and allow to .stand on
the steam bath for 20-30 nmuiies. .Allow tliessilutiun to stand at least 1 hour and
cool to room temperature during this rime Filter otVltie precipitate through a
weighed filtering crucible; test the filtrate for complete precipitation with a little
dimethylglyo.xime solution, and wash the precipitate with cold water until free
from cliloride. Dry the precipitate at 100-120 for 45 - 60 minutes, and weigh as
NilQfDOjN,!..

Calculate the pcrcciUage of nickel in the steel.

For other methods for the determination of nickel, see .Section XI, 47.

* Alieni.uucJ>. suiiit-icnl of ,i nicJ,-cl sjit lo conum .sKiut U 0 j -O.O.s ^ of iiicUl may tv u,'cd.

t Bureau of .Viial) vcd S.mipics •Nickct .Sled. No iyy (a tliiush Clicwical .SlaiiCardl is suilublc. This
sled conlains .iKiul 3 5 (vr cem of mcKd.

, II a prccipit,itc appc.irs irr il ihc sidulioii is lull dear islicti it is reiuIcrcU amilioiiiacal. uiorc lartaric
or alric .ici J mim be aildea uiuil .i (vifcsilj clear solulioii is ohlaiiie,! upon aildiiig dibile aiiuiwnia
solution An> insoluble mailer slanilil be lilietcU iiir,inU waslieJ uith hot waler conlaiiiing a little
afniiumiit Axiluuon,

448

GRAVIMETRY XI, 25/26

Systematic gravimetric analysis

XI, 25. GENERAL DISCUSSION. In the succeeding sections a brief
account will be given of a. number of selected methods for the gravimetric
determination of the various elements and radicals. It is believed that these will
suffice to meet the needs of the student during the whole period of his training;
for a more detailed study, particularly of the limitations of some of the various
methods, reference must be made to other treatises (see Selected Bibliography at
end of this chapter).

As no arrangement of ions or metals is ideal the following procedures have
been arranged in alphabetical order with the metals (cations) listed first and the
anions similarly arranged afterwards.

Cations

XI, 26. ALUMINIUM. Methods for the determination of aluminium as the
oxide and as the 8-hydroxyquinolate have already been given (Sections XI, 18
and 19); the following procedure, which also involves ignition to aluminium
oxide, is also widely used.

Determination of aluminium as basic aluminium succinate and subsequent
ignition to the oxide, AljOg (precipitation from homogeneous solution).
Discussion. Aluminium can be precipitated from homogeneous solution as the
dense basic succinate by boiling an acid solution containing succinic acid with
urea (starting pH = 3. 1-3. 5). The hydrolysis of urea gradually ' produces
ammonia, resulting in a pH of 4.2-4.6 :

C0(NH2)2 + H2O = CO2 + 2NH3

The dense precipitate is easily filtered and washed, and exhibits much less
tendency to adsorption of other salts than does the precipitate obtained by
precipitation as the hydroxide. Upon ignition, the basic succinate is readily
converted into aluminium(III) oxide.

The method permits the separation of aluminium from large amounts of
calcium, barium, magnesium, manganese, or cadmium, or from unequal amount
of nickel or cobalt ; for large amounts of nickel and cobalt, a double precipitation
is necessary. Owing to the relatively low solubility of copper(II) succinate, the
copper(II) must first be reduced by hydroxylammonium chloride solution or
ammonium hydrogensulphite solution at the boiling point. A double pre-
cipitation is essential if zinc is present. Iron(III) must be reduced to iron(II) : the
hot hydrochloric acid solution containing the aluminium sample is first reduced
with fresh ammonium sulphite solution, precipitation is then effected in the
presence of 2 cm^ of phenylhydrazine, and the precipitate is washed with 1 per
cent succinic acid solution (containing some phenylhydrazine) rendered neutral
to methyl red with aqueous ammonia.

Procedure. The solution should contain about 0. 1 g of A1 and be acid with
hydrochloric acid. Add dilute ammonia solution until the solution becomes
slightly turbid, remove the turbidity with dilute hydrochloric acid, and add 1-2
rops in excess. Add a solution of 5 g of A.R. succinic acid in 100 cm^ of water,
0 lowed by 10 g of ammonium chloride and 4g of urea; dilute to 500 cm^ with
istilled water. Heat the solution to gentle boiling and continue the boiling for 2

449

XI, 27/28 quantitative INORGANIC ANALYSIS

hours after the solution has become turbid (ca. 45 minutes).* The insertion of a
boilint; rod into the solution is recommended as tliis reduces the tendency to
‘bump’ during the heating. Allow the precipitate to settle for a few minutes, filter
through a Whatman No. 40 or No. 540 filter paper, and wash several times with a
1 per cent succinic acid solution made neutral to methyl red with aqueous
ammonia solution. If any precipitate adheres to the sides of the beaker, dissolveit
in a little dilute hydrochloric acitl, add a drop of methyl red or phenol red
indicator, and then dilute ammonia .solution until just alkaline: filter olT the
precipitate of aluminium hydro.vide on a separate small filler paper, and wash it
with a 2 per cent solution of ammonium nitrate. Place both papers and
precipitates in a silica or, preferably, a platinum crucible, and ignite to constant
weight at 1200 C (compare .Section XI. 18). Weigh as .AUOj.

XI, 27. A.MMO.NIU.M. Dhamion. I'or the determination of ammonium
by a gravimetric procedure, it must be present as the chloride; all other cations
must be absent. A little h.wlrochloric acid is added, followed by e.xcess of
chloroplaiinic acid reagent (sec Xcciion .\1, 50C). The mixture is evaporated
almost to dryness on the water bath; the residue is triturated with absolute
ethanol to remose excess of cliluroplatinic acid, and then Iransfcned to a
weighed filtering crucible (Gooch, sintered glass, or porcelain). The crucible is
dried at 1.^0 C. and tlie residual ainmonium chloroplatinate, {NH^)jPtCl 4 ,
weighed. (For details of experimental procedure, see Section XI, 50C.)

Ammonium may uLo be determined by precipit.ition with sodium
tetraphenylborale as the sparingly soluble armnonluin letraphenylboron
NH 4 {U(QHj) 4 ]. using a similar priwcduie to that described for potassium; it is
dried at U)0 C. For further details of the reagent, including interferences,
notably potas.sitim, rubidium, and caesium, .see Section XI, 50C.)

if tile ammonuiin salt is present with other cations and anions, a titriineiric
procedure (see Chapter X) is usually employed.

XI, 28. ANTI.MO.NV. .Antimony may be determined in the following forms;

.V. Antimony (III) sulphide, SbjSj. Dticus.uvn. Tliri method is of limited
application, since no other c!cmeni.s that are precipitable by hy drogen sulphide in
acid solution can be present, and tiic sulphide must be dried and finally heated in
an atmosphere of carbon dioxide at 2H0'.1U0 C. .Arsenic can be separated by
removal by distillation as arsenic trichloride; tin c.m be remosed by precipitation
in the presence of oxalic and l.irtaric .icids or of phosphoric acid.

PruccJurc. Quickly heat the solution of the antimony compound in 114-
hydrochloric acid (100 cm') (1) contained in a conical liask to boiling and
immediately pass a rapid stream of washed hydrogen sulphide; maintain the
solution at 90- 100 C. .Sluike the llask gently at interxals after the sulphide has
tunicd red, and keep the precipitate, as far as possible, bcKnv the surface of the
solution. As the precipitate darkens in colour, reduce the ga.s stream. Continue
the pas.sage of gas until the precipitate is crystalline and black in colour (totitl
time requircrl for precipitation i.s .10-35 ininute.s). Dilute the .solution witli an
equal volume oi water, mix. and heat again whilst the gas is slowly passed into the

1 he boihny period (.liter ihc .ipjK.ir.iiicc or* a turbidity) riuy be roluecd lo 1 hour by first pjrtiail)
ncuirjiiiing i he hot ioluium to brotno-pheno! blue or toiv«cih)l by ihcdrop-^vixc

dUutc am iiionu solution; a very lainl opalcvccncc will appear.

450

GRAVIMETRY XI, 29

suspension for some minutes. When the solution is clear, cool, and filter through
a filtering crucible (Gooch, sintered glass, or porcelain) that has been heated at
280-300 °C and weighed. Wash the precipitate a few times with water to rernove
acid, and then with ethanol, draw air through the crucible to dry the precipitate
as far as possible. Place the crucible and contents in a wide glass tube passing
through an electrically heated tube furnace. Heat for 2 hours at 100-130 °C in a
current of carbon dioxide (this will completely dry the precipitate), and then heat
for a further 2 hours at 280-300 °C (this process will convert any Sb2S5 present
into SbjSj and will volatilise the sulphur). Cool in a slow stream of carbon
dioxide, then place in a desiccator for 20—30 minutes, and weigh as Sb2S3.

Note. 1 . A solution, suitable for practice in this determination, may be
prepared by dissolving 0.5 g, accurately weighed, of A.R. anhydrous antimony
potassium tartrate in 1 50 cm^ of 1 : 4-hydrochloric acid.

B. Antimony pyrogaUate, Sb(CfiH503). Antimony(III) salts in the pre-
sence of tartrate ions may be quantitatively precipitated with a large excess of
aqueous pyrogallol as the dense antimony pyrogallate. The method allows of a
simple separation from arsenic; the latter element may be determined in the
filtrate from the precipitation of antimony by direct treatment with hydrogen
sulphide.

Procedure. The solution should contain the antimony (0. 1-0.2 g) in the
trivalent condition. Add a slight excess over the calculated quantity of potassium
sodium tartrate to avoid the formation of basic salts upon dilution. Dissolve
approximately five times the theoretical quantity of pure pyrogallol (Section XI,
IIN) in lOOcm^ of air-free water, add this all at once to the antimony solution,
and dilute to 250 cm^. After 30-60 seconds the clear mixture becomes turbid, and
then a dense, cloudy precipitate forms which separates out rapidly. Allow to
stand for 2 hours, filter through a weighed sintered glass or porcelain filtering
crucible, wash several times with cold water to remove the excess of pyrogallol
(50 cm^ is usually sufficient), dry at 100-105 °C to steady weight. Wash again
with cold water, dry at 100-105 °C, and weigh; repeat the operation until the
weight is constant. Weigh as Sb(C6H503).

It should be pointed out that the titrimetric methods described for the
determination of antimony (Chapter X) are to be preferred to the gravimetric
methods as they are simpler, more rapid, and quite as accurate.

XI, 29. ARSENIC. Arsenic may be determined in the following forms-

A. Arsenic(III) sulphide, AS2S3. Discussion. The arsenic must be present
in the trivalent state. Arsenic in the trivalent state (ensured by the addition of, for
example, iron(II) sulphate, copper(I) chloride, pyrogallol, or phosphorous acid)
may be separated from other elements by distillation from a hydrochloric acid
solution, the temperature of the vapour being held below 108 °C; arsenic
trichloride (also germanium chloride, if present) volatilises and is collected in
water or in hydrochloric acid.

Procedure. Pass a rapid stream of washed hydrogen sulphide through a
solution of the arsenic(III) (1) in 9M-hydrochloric acid at 15-20 °C. Allow to
stand for an hour or two, and filter through a weighed filtering crucible (Gooch
sintered glass, or porcelain) (2). Wash the precipitate with 8M-hydrochloric acid
saturated with hydrogen sulphide, then' successively with ethanol carbon
disulphide (to remove any free sulphur which may be present), and ethanol Dry

at 105 °C to constant weight, and weigh as AS 2 S 3 .

451

XI, 30 QUANTITATIVE INORGANIC ANALYSIS

Notes. 1. A suitable solution for practiee in this dciennination is prepared
by dissolving about 0.3g of A.R, arsciiic(Jjll) oxide, aecuralely weighed, in 9,1/-
hydrochJoric acid.

2. Soinciiines a film of the siilpliide adheres to the glass vessel in which
precipitation was carried out; tliis can be dissolved in a little ammonia solution
and the sulphide rc-precipitatcd with the acid washing litiuor.

B, Ainnioniuiu uranyl arsenate, Nn 4 U 02 As 04 ,.rHj 0 , and subsequent
weigtiinK as the o.xide. UjO». Tlic addition of a urany! salt .solution to an
itiH'iuiU' solution containing cxce.ss of ammonium ion.s resulLs in the pre-
cipitation of ammonium uranyl ar.senatc, which is soluble in mineral acid.s but
insoluble in acetic acid. Upon igniting the precipitate, the ar>;enic is completely
volatilised, Ic.iving a moss-green residue which consi.st.s mainly of UjO^: this
residue is dissolved in concentrated nittic acid, and the resultant uranyl nitrate
upon cautious evaporation and ,sub:.equcnt ignition yields pure black UjO^, and
is weighed in this form.

If tiie solution contains arscnitc. the latter must iirst be o.\idi.sed to arsenate
with 0, 1 A'-polassium brom.ilc m Itydrochloric acid .solution at 70 C in the u-sual
way {Section X, 139). .A method i.s thus .ivailablc for the determination of arsenite
and arsenate in admixture. The arscnitc is tirsi determined with standard
potassium bromate solution, and tlie tola! aisenatc in the resulting liquid is then
determined by prccipit.ition as the unmium salt.

Proii'iliirc. The solution (ISOcm^) siunild contain about O.Obg of .-ks as
arsenate. Add .lOciif' of 4,\/-aminonta .solution, acidify with acetic acid, heat to
boiling, and add SO cm’ texcessS of .tppr\>ximaic!y O.t.V-uranyl acetate solution.
Allow to stand for several hours, hut preferably overnight ; during this period the
pale-yellow gr.uni!ar ptccipilate will become coarser. Filter through a fine
quantitative filter paper, wash free from soluble salts, and transfer the filter and
precipitate to a wcigiied silica crucible, //cut in a j\tnw cluunhcr pnnhk'il wilh a
{’oini iltau^ht until all the catbun has burnt oil' - the arsenic is simultaneously
volatilised. Moisten the residue with u few drops ofconcentraied nitric acid, and
Ignite to constant weight over an ordinary Bunsen burner. Weigh as UjOj.

.XI, 30. B-AKIUM. The various methods available for the determination of
barium all siilfcr from the disadvantage that they are aiTected by a number of
intericring metals. 7 wo main procedures arc commonly used:

A. Dflcrmirialiou of barium jus sulphate, /iuc.'isimn. This method is most
w'idcly employed. The clfect of various imerfering elements and radicals (c.g.,
calcium, strontium, lead, nitrate, etc., which contaminate tlie precipitate) is fully
dealt with in Section .XI, H4- Ihe solubility of hitrium sulphate is cu. 1 part in
400000 ol cold water or about 2.5 mg dm" T The solubility is greater in hot water
or in dilute hydrochloric or nitric acid, and less m soUuion.scomuiiung a common
ion.

The barium sulphate may be precipitated cither by the useufsulphuricacid. or
Iron) homogeneous solution by the use of sulphamic acid solution which
produces sulphate ions on boiling;

NH,SO,,H4-li-0 a-SO.i-'+fr

Proa’dure. Precipitation with sulphuric acid. The solution (lOOcin^)
should Contain not more than O.I5gof barium (I ). and not more than I percent

452

GRAVIMETRY XI, 30

by volume of concentrated hydrochloric acid. Heat to boiling, add a slight excess
of hot O.SAf-sulphuric acid slowly and with constant stirring. Digest on the steam
bath until the precipitate has settled, filter, wash with hot water containing two
drops of sulphuric acid per litre,' and then with a little water until the acid is
removed. Full experimental details of the filtration, washing, and ignition

processes (900-1000 °C) are given in Section XI, 84. Weigh as BaS 04 .

Note. 1. A suitable solution for practice may be prepared by dissolving
about 0.3 g, accurately weighed, A.R. barium chloride in 100 cm^ water and
adding 1 cm^ of concentrated hydrochloric acid.

Procedure. Precipitation from homogeneous solution; sulphamic acid
method. The sample solution may contain up to 100 mg of barium, preferably
present as the chloride. A solution prepared from about 0.18 g, accurately
weighed, A.R. barium chloride may be used to obtain experience in the
determination. Dilute the solution to about 100 cm^ ; add 1 .0 g sulphamic acid.
Heat the covered beaker on an electric hot plate at 97-98 °C ; continue the heating
for 30 minutes after the first turbidity appears. Filter through a weighed
porcelain filtering crucible and wash with warm distilled water. Ignite to constant
weight at 900 °C (preferably in an electric muffle furnace). Weigh as BaS 04 .

B. Determination of barium as chromate. This method is of limited
application because of the influence of numerous interfering elements. It is
useful, however, in the separation of barium from both calcium and strontium.
Thermogravimetric analysis suggests that a drying temperature of 120-1 80 °C is
to be preferred to the much higher temperatures usually recommended; the loss
of oxygen amounts to about 1 per cent at 1000 °C.

Precipitation of barium chromate is usually carried out in a dilute acetic acid
solution which is buffered with ammonium acetate; a double precipitation is
desirable in the presence of much strontium and/or calcium.

Barium chromate is soluble in an acid solution of pH about 2. If the pH of such
a solution is slowly raised by the use of urea the barium chromate will precipitate
in large, readily filterable crystals ; ammonium acetate is added to prevent the pH
increasing too rapidly. For quantitative precipitation the final pH should be near
5.7. Above pH 5.7, strontium will precipitate if present in large quantities.
Barium (100 mg) may be separated satisfactorily from calcium (100 mg) and
strontium (40 mg) in a single precipitation; for larger quantities of strontium
(50-100 mg) a double precipitation is necessary. If strontium is absent the final
pH may be 6.8-7.0. , ,

Procedure. The solution (200 cm^) ( 1) should contain not more, than 0.4 g
of Ba, and be neutral in reaction. Add 1.0 cm^ 6Af-acetic acid and lOcm^ neutral
3M-ammonium acetate to the solution, heat to boiling, and, treat with a slight
excess of a hot dilute solution of ammonium chromate (2) which is added
dropwise from a burette with constant stirring. Place the beaker on a water bath
until the precipitate settles; test for completeness- of precipitation by adding a
little more of the reagent. Allow to cool, filter through a weighed porcelain or
sintered glass filtering crucible, wash with hot water until 1 cm^ of the washings
gives scarcely any reddish-brown coloration with neutral silver nitrate solution.
Dry to constant weight at 120 °C. Weigh as BaCr 04 .

Notes. 1. A suitable solution for practice in this determination may be
prepared by dissolving about 0.3 g, accurately weighed, A.R. barium chloride in
200 cm^ of water.

2. The ammonium chromate solution is prepared by dissolving 10 g pure

453

XI, 31 QUANTITATIVE INORGANIC ANALYSIS

ammonium dichromatc (free from sulphate) in lOOcm^ water, and adding dilute
ammonia solution until the colour oi the solution is clciir yellow.

XI, 31. BIsRYl.LlUM. Beryllium may be determined in the following
forms:*

A. Determination of beryllium by precipitation with ammonia solution and
subsequent ignition to beryllium oxide. Dixatsuon. Beryllium may be de-
termined by precipitation with aqueous ammonia solution in the presence of
ammonium’ chloride or nitrate, and subse<iucntly igniting and weighing as the
oxide BeO. The method is not entirely s.itisfactory ow ing to the gelatinous nature
of the precipitate, its iciideney to adhere to tiie sides of the sessei, and the
possibility of adsorption elfecis.

Beryllium is sometimes precipitated together with aluminium hydroxide,
which’it resembles in many respects. Separation from aluminium (and rdso from
iron) may be elfected by means of oxinc, .Xn acetic acid solution containing
ammonium acetate is used; the aluminium and iron arc precipitated a.soxinate5,
and the beryllium in the tilirate is then precipitated witli ammonia solution.
Phosphate must be absent in the initial precipitation of bciy ilium and alunnnium
hydroxides.

The precipitation by ammonia solution of such clemcnu as AI, Bi, Cd, Cr, Ca,
Cu. Fc, Pb, Mn. Ni. and Zn m.iy be prevented by comple.xation with FDTA:
upon boiling the ainmoniacal solution, beryllium hydroxide i.s precipitated
quantitatively.

In all the above methods ilic element is weighed as the oxide, BeO, which is
.somewhat hygroscopic (compare alumimumdl!) oxide). The ignited residue,
contained in a covered crucible, must be cooled in a desiccator contaim'ng
concentrated sulphuric acid or phosphoras penioxidc. and weighed immediately
it has acquired the laboratory tempcr.itute,

Procaliirc. The beiyllium solution (ZOUenr*), prepared with nitric acid or
hydrochloric acid and coiuaining .ibout 0. 1 g of Be, mu.sl be almost neutral and
contain no other substance prccipitable by ammonia solution. Heal to boiling,
and add dilute mnmonia .solution slowly and with constant stirring until present
in is ri- s/ig/i/ excess. Add a Whatman accelerator or onc-half of a Whatraau
ashlcss tablet, boil for 1 or 2 minutes. ,md filler on a Whatman No. 41 or 541 filter
paper. I ranster as much of the precipitate as possible by rinsing with hot 2 per
cent ammonium nitrate solution. Remove any precipitate, adhering to the wails
ol the beaker by dissolving in ilic inmiiiuim volume of liot very dilute nitric acid,
heating to boiling, and precipitating as before. Filter through the .same paper,
and wash thoroughly with the ammonium nitrate solution. Place the paper and
precipitate in a weighed silica or platinum crucible, dry. and then slowly
decompose the hydroxide by raising the temperature gradually to 70Q’C, and
then ignite at about i()i)0 "C for at fea.sl 1 hour. Cool in a covered crucible in a
desiccator charged vviih concenlratexl sulphuric acid, phosphorus pctiloxide, or
anhydrous magnesium percliloratc, and weigii immcdi.ilcly when cold as BeO.

In tile presence of uiterfcriii” clement.s, proceed as follows. Neutralise
S012()cm'* of the solution containing 15-2,5 mg of beryllium vviiii ammonia

llcr>llium anil its conipuiiniK arc loxic .iinl eatc ihuutil I'c l.tkcii lo .ooid inhatalien of dusU or
conlaclwilh ciciand ikiii.

454

: GRAVIMETRY XI, 32

solution until the hydroxides commence to precipitate. Redissolve the precipitate
by the addition of a few drops of dilute hydrochloric acid. Add 0.5 g of
ammonium chloride and sufficient 0.5M-EDTA solution to complex all the
heavy elements present. Add a slight excess of dilute ammonia solution, with
stirring, boil for 2-3 minutes, add a little ashless filter pulp, filter, and complete
the determination as above. •

B. Ammonium beryllium phosphate and subsequent ignition to beryllium
pyrophosphate. Discussion.’ Beryllium may be precipitated as ammonium
beryllium phosphate and subsequently ignited to the pyrophosphate, Be2P207.
If the- conditions of precipitation are not carefully controlled, slight departures
from the expected theoretical Be2P207 composition may occur, resulting in loss
of accuracy. In the presence of aslight excess of EDTA, metal ions such as those of
aluminium, iron, copper, nickel, calcium, and magnesium heed not be removed.

Procedure. Precipitate the hydroxides, containing from 1 to 10 mg of
beryllium, with ammonia solution at pH 8-9. Filter off the precipitate, transfer
back to the original beaker, and dissolve it in dilute mineral acid. Dilute the
solution to lOOcm^ and adjust the pH to 2. AddScm^ 15 per cent diammonium
hydrogen phosphate solution and a slight excess of 15 per cent EDTA solution
(according to the amount of interfering elements present) : both reagent solutions
should be previously adjusted to a pH of 5.5. Now add 0.5M-ammonium acetate
to the resulting solution until the pH is 5.5 (use a pH meter). Digest the solution
just below the boihng point for 5-1 0 minutes, cool, filter the granular precipitate,
redissolve in the minimum volume of hot 6M-hydrochloric acid, and repre-
cipitate, using only 1 cm^ of each of the reagents. Filter again, wash with a 0.5M-
acetate buffer (3.5 g ammonium acetate and 3.0 cm^ glacial acetic acid per
100 cm^ of water), ignite at 1000 °C, and weigh as Be2P20.7 in the usual manner.

XI, 32. BISMUTH. Bismuth may be satisfactorily determined in the follow-
ing forms :

A. Determination of bismuth as oxyiodide. Discussion. The cold bismuth
solution, weakly acid with nitric acid, is treated with an excess of potassium
iodide when Bila and some K[Bil4] are formed :

Bi(N03)3 + 3KI = BiU + 3KNO3
Bil3 + KI = K[Bil4]

Upon dilution and boiling, bismuth oxyiodide is formed, and is weighed as such
after suitable drying.

Bil3 +H2O = BiOI+2HI

(black)

K[Bil4] +H2O = BiOI + KI -t- 2HI

(yellow)

A large excess of potassium iodide should be avoided, since the complex salt is
not so readily hydrolysed as the tri-iodide. This is an excellent method, because
the oxyiodide is precipitated in a form which is very convenient for filtration and
weighing.

Procedure. The cold bismuth nitrate solution, containing 0. 1-0. 1 5 g of Bi
(1), must be slightly acid with nitric acid (2), and occupy a volume of about
20 cm . Add finely powdered solid potassium iodide, slowly and with stirring

455

Xl, 33 QUANTITATIVE INORGANIC ANALYSIS

unlil the supernatant liciuicl above the black precipitate of bismuth tri-iodide is
just coloured vcilow (due to KlUiUj). Dilute to 2()0cni^ with boiliiit- water, and
boil lor a lew minutes. The black iri-iodide is converted into the copper-coloured
precipitate of the oxyiodide. The supernatant liquid should be colourless; if this
i.s yellow, a further 100 cm* of water should be added, and the boiling continued
until colourless. Add a few drops of methyl orange indicator, and then sodium
acetate .solution (25 g dm ' from a buietic until the solution is neutral. Filter olT
the precipitate through a vseighed CJooch. sintered glass, or porcelain filtering
crucible, wash with iiot water, and dry at t05--l 10 C to coivstant weight. Weigh
as iJiOI.

Notes. 1. A suitable solution for pt.icticc c.in lx: obtained by dissolving
about 0.15 g of pure bi.smuth. accurately neighed, in the minimum volume of 1 ;4
nitric acid. Alteruaiively, A-K. bismuth nitrate may be used,

2. Chloride and bromide should be absent. If the solution is strongly acid with
nitric and. it should be cvai>or.itetl to dryness on the water bath, and the residue
dissolved in a little dilute nitric acid.

U. Uetenuinalion of hisntuth as pynigallate. DiMuxuun. The pre-
cipitation of bi.smutli with pjtogallol is quaniiiati'-c only if the acidity
(hydrochloric, sulphuric, or mtnc acid 1 d^ics not exceed (I. I .V. Tlie method is an
e.xccllcnt one for the dcicrmination of bisimnli in the presence of lead, cadmium,
and zinc. Antimony, which forms a -similar complc.x. mu.st, of course, be ahsciit.

Pnhcdiirc. The solution (15l>cm') should be weakly acid with nitric acid
and contain b. 1 -0.2 g of Hi, Tie.it the solution with dilute ammonia solution until
a permanent luibidiiy is obtained; tender the solution clear by the cautious
addition of a little dilute nitric acid, llcat to boiling, and add a .slight e.vccssofa
solution of pure pyrogalloi (Section Xl, I IN) m air-free water, A yellow, fmdy
crystalline precipitate is miiiicdi.itely formed, iiuil for a short time, test for
completencs-s of precipitation with a little of the reagent, diluieslighlly. .ind iillcr
through a weighed sintered glass or porcelain lilteiing crucible after the
precipitate settles out. W.ish with U.t)5.U-niiiic acid, and finally with water. Dry
toconstant weight at 105 C. Weigh as BitQfi^Ou.

The following alternative mciiiod has been iccommeiukd. .Add l.Ogorpiirc
pyrogalloi to the solution (liUcin*) containing about 0.1 g of Bi and heated to
70 C. Then add 0,5.U-aqucous ammonia solution dropwisc until a distinct
turbidity forms. Meat the resulting solunou to boiling, add 2 drops of thymol
blue indicator, and then more of the ammonia solution until the solution is basic.
Heal on a water bath fiir 1 0 minutes, lilter on a sintered gla.s.s or porcelain filtering
crucible, wash, and dry to constant weight at 105 C. Wkigh as Bi(QHaOj).

XI, 33. CAD.MIUM. Cadmium may Ik determined in the following forms:

A. Determination of cadmium as the 2-na[)htha(iuinoline complex,
Discus.\km. This method [Kriuits of the separation of c.idniium from a-la-
lively large quantities of zinc, iron, chromium, alumitiium, cobalt, nickel,
manganese, and magnesium, and also from antimony and tin if ammonium
oxalate or large aiiunmis of sodium tartrate are u.scd.

Procedtiri!. The c-.idmiunj salt solution, containing about 0.15 g ol Cd,
should occupy a volume of about 50 cm^ and be M with respect to sulphuric
acid. Add SOciir' of 10 per cent sodium tartrate solution, followed successively

•156

GRAVIMETRY XI, 33

by a 2.5 per cent solution of 2 -naphthaquinoline in 0.25M-sulphuric acid, a few
drops of dilute sulphuric acid, and then 0.2M-potassium iodide in excess. After
20 minutes, filter the precipitate of the cadmium complex through a weighed
Gooch, sintered glass, or . porcelain filtering crucible, wash with a solution
containing lOcm^ of 0.2M-potassium iodide, lOcm^ of 2.5 per cent 2-
naphthaquinoline in 0.25Af-sulphuric acid, 80 cm^ of water, and 1-2 drops of
dilute sulphur dioxide solution, and finally suck as free as possible from the wash
liquor. Dry the precipitate to constant weight at 130°C. Weigh as

[(Cl3^9N)2H2](Cdl4). -J

B. Determination of cadmium as quinaldate. Discussion. Qmnaldic acid
or its sodium salt precipitates cadmium quantitatively from acetic acid or neutral
solutions. The precipitate is collected on a Gooch type of crucible, and dried at
125 °C. A determination may be completed, in about 90 minutes. For the
limitations of the method, see Section XI, HM.

Procedure. The solution (150 cm^) should be neutral or weakly, acid with
acetic acid, and should contain 0. 1-0. 1 5 g of Cd. Heat the solution to boiling, and
remove the source of heat. Add the reagent (a 3.3 per cent solution of quinaldic
acid or of the sodium salt in water) dropwise with vigorous stirring until present
in slight excess. Then neutralise carefully with dilute ammonia solution, and
allow the white curdy precipitate to settle. When cold, wash with cold water by
decantation, filter through a sintered glass or porcelain filtering crucible, wash
thoroughly with cold water, and dry at 125 °C to constant weight. Weigh as
CdfCioHeO^N)^.

C. Determination of cadmium by the pyridine method. Discussion. If a
hot neutral or faintly acid solution of a cadmium salt is treated with ammonium
thiocyanate and pyridine, dipyridinecadmium thiocyanate is quantitatively
precipitated. This precipitate is collected, and washed, inter alia, with ethanol
and diethyl ether containing a little pyridine; it may be dried simply by leaving in
a vacuum desiccator for 15-20 minutes. A determination can thus be completed
in less than an hour. If the solution is weakly acid, ammonium thiocyanate may
be added, followed by pyridine, until a precipitate just forms, the latter dissolved
by warming, and a further 1 cm^ of pyridine added.

Procedure. The solution (75-100 cm^) should contain about 0. 1 g of Cd (1 )
and be neutral or very feebly acid. Add 0.5-1. 0 g A.R. ammonium thiocyanate,
stir, heat to boiling, and treat the solution with 1 cm^ pure pyridine dropwise and
with stirring. The complex slowly separates as the solution cools. Filter the cold
solution through a weighed sintered glass or porcelain filtering crucible, transfer
the precipitate to the crucible with the aid of Solution 7.. Wash four to five times
with Solution 2, then twice with 1 -cm^ portions of Solution 3, and finally five to
six times with small volumes {ca. 1 cm^) of Solution 4. (For further experimental
details, see under Zinc, Section XI, 62.) Dry the precipitate in a vacuum desiccator
(Fig. Ill, 15) for 10-15 minutes and weigh. Repeat the drying until constant
weight is attained. Weigh as [Cd(C 5 H 5 N) 2 ](SCN) 2 .

Solution 1. 100 cm^ water containing 0.3 g NH^SCN and 0.5 cm^ pyridine.

Solution 2. 73 cm^ water, 25 cm^ 95 percent ethanol, 0 . 1 g NH 4 SCN, and 2 cm^
of pyridine.

>So/m/oni. 10 cm^ absolute ethanol and 1 cm^ pyridine. . . .

Solution 4. 1 5 cm^ diethyl ether (sodium dried) and 2 drops pyridine.

Note. 1 . For practice in this determination use about 0.3 g, accurately

weighed, A.R. cadmium sulphate or A.R. cadmium iodide. ’

457

XI, 34/35 QUANTITATIVE INORGANIC ANALYSIS

XI, 34. CALCIUM. Of (he methods available for the determination of
calcium, that via calcium carbonate by initial precipitation of the oxalate is the
best and most widely used.

A. Determination of calcium as calciunr carbonate, by precipitation from
lioniogcneous solution as the oxalate. Dhamimt. T'hi.s method h;u> been fully
described in Section XL 20. Precipitation is first ellceted as calcium oxalate which
is subsequently converted into calcium carbonate or calcium o.xide.

Precipitation of calcium oxalate may also be made from bomogeneous solution
either by u.sc of urea as reaeeiU (sec Section X, 94) tif by the use of dimethyl
oxniate. Hoth procedures lead to satisfactory separations from magnesium.

Procalure. To obtain experience in tiiis method, weigh out accurately
about 0.25 g A.R. calcium carbonate, dissolve it in 5 cin^ dilute hydrochloric acid
(1 ;5), and dilute to 150cm ‘. .Adjust the pH to 4.7 with dilute ammonia solution
(use a pH meter). Add lOOem^ ammonium acetate .icetic acid bulTer (2.53/ with
re.ipect to each) and 5.0g pure dimethyl oxalate ( ! ). Cover the beaker and heat on
a temperature controlled hot plate at 90 C for 2.5 hours: stir cKcasionaily.
Precipitation u-sually commences after 10 minutes. As a precautionary measure
add. 10 minutes before liltration, Sem-* t>f a solution containing 0.25g
ammonium oxalate. Coo! the solution r.ipidly to room lemjveraiure. filter
through a weighed porcelain filtering crucible of medium porosity, and wash
with 1 per cent ammonium ox.)latc solution. Dry the precipitate for 1 hour at
1 20 C and then ignite in an electric imitlle furnace at 500 C for 2 hours, Weieh as
CaCOj.

Note. 1 , Unless pure dimethyl oxalate is used, immediate precipiiaiion of
some line calcium oxalate will occur. Impute dimethyl oxalate should be
reery.slalli.sed from ethanol and stored m a desiccator.

U. Determination of calciutu as tungstate. Discussion. The calcium is
precipitated in neutral solution tpH 7-8) with a solution of sodium tungstate .is
calcium tungstate CaWO^. This piiKcdurc is applicable in the presence of
considerable amounts of magnesium.

Procedure. The solution (!0()cm“) should contain about 0.04g Ca and
possess a pll of 7-8 (1). Add either dilute sodium hydroxide solution or dilute
acetic acid to att.iin the correct pH with the aid of crcsol red indicatoror a pH
meter. Large quantities of .munomum sah.s hinder precipitation. He.it the
solution to about 80 C and introduce, with stirring, 2.()cnr' of the sodium
tungstate reagent (2). Calcium tungstate is precipitated immediately. Cool in ice
for 30 minutes, filter through a sintered glass (porosity No. 5) or porcelain
filtering crucible, wash with 20etu-* warm water, and dry at 1 10 C for I hour or
to constant weight. Weigh as CaW'04.

■Notes. 1 . For practice in this determination u .solution may be prepared by
dissolving l.Og (accurately weighed) A.R. calcium carbonate in a little dilute
hydrochloric acid, and diluting to 250 cm-* in a graduated fiask. A 25-cin^ portion
ol this solution, diluted to lOOeni'’, may be used.

2. Tile reagent is prepared by dissolvine 19.0 g .A.R. sodium tungstate,
Naj\V04,2H20, in lOOem^ of water.

XI, 35. CLRIUM. Determinaliun of cerium a.s ceriuni(lV) iodate and
subsequent ignition to ccriutnflV) oxide. Discussion. Cerium may be de-
termined as ceriuni(lV’) iodate, CeflOj)^, vvhicli is ignited to, and weighed as, the
o.xide, CeOi- Thorium (al.so litaniuin and /.irconiutn) must, however, be first
removed (sec Section XI, 56); the method is then applicable in the presence of

458

GRAVIMETRY XI, 36

relatively large quantities of rare earths.' Titrimetric methods (see Section X, 113)

are generally preferred. •

Procedure. The solution should not exceed 50 cm^ in volume, all metallic
elements should be present as nitrates, and the cerium content should not exceed
0. 10 g. Treat the solution with half its volume of concentrated nitric acid, and add
0.5 g potassium bromate (to oxidise the cerium). When the latter has dissolved,
add ten to fifteen times the theoretical quantity of potassium iodate in nitric acid
solution (1) slowly and with constant stirring, and allow the precipitated
cerium(IV) iodate to settle. When cold, filter the precipitate through a fine filter
paper (e.g., Whatman No. 42 or 542), allow to drain, rinse, once, and then wash
back into the beaker in which precipitation took place by means of a solution
containing 0.8 g potassium iodate and 5 cm^ concentrated nitric acid in 100 cm^.
Mix thoroughly, collect the precipitate on the same paper, drain, wash back into
the beaker with hot water, boil, and treat at once with concentrated nitric acid
dropwise until the precipitate just dissolves (20-25 cm^ of acid are required per
0.1 g of cerium). Add 0.25 g potassium bromate and as much potassium iodate-
nitric acid solution as before. When cold, collect the cerium(IV) iodate upon the
same filter paper, wash once with the washing solution, return to the beaker, stir
with the washing solution, filter again, and wash thrice with the same solution.
Place the filter paper and precipitate in the same beaker, add 5-8 g oxalic acid
and 50 cm^ water, and heat to boiling. After all the iodine has been expelled, set
aside for several hours, filter, wash with cold water, dry, and ignite (at 500-600 °C)
to constant weight in a platinum crucible. Weigh as CeOj.

Note. 1. This is prepared by dissolving 50 g of potassium iodate in
167 cm^ of concentrated nitric acid, and diluting to 500 cm^.

XI, 36. CHROMIUM. Chromium may be determined in one of the following
forms:

A. Determination of chromium as barium chromate. Discussion. The
chromium must be present as chromate. The method is of limited application
because of the general insolubility of chromates. Chlorides do not interfere, but
sulphates must, of course, be absent. For further properties of barium chromate,
see Section XI, 30B.

To convert a chromium(III) salt into a chromate, treat the chromium solution
contained in a porcelain dish with several cm^ of bromine water, followed by
freshly prepared potassium hydroxide solution until alkaline. Warm until the
odour of bromine disappears.

Procedure. The solution should contain about 0. 1 g of Cr as chromate, be
neutral or weakly acid with acetic acid, and occupy a volume of 200-300 cm^.
Add a 10 per cent solution of barium acetate dropwise from a burette and with
constant stirring to the boiling solution (1) until present in slight excess. Place the
beaker on a water bath until the precipitate settles; test for completeness of
precipitation by adding a little more of the reagent. Allow to cool, filter with
gentle suction through a weighed Gooch, sintered glass, or porcelain filtering
crucible, wash with hot water until 1 cm^ of the washings gives no precipitate
with a little dilute sulphuric acid. Complete the determination as described in
Section XI, 30B. Weigh as BaCr 04 .

Note. 1 . A solution for practice may be prepared by dissolving about 0.5 g

A.R. potassium dichromate, accurately weighed, in 300 cm^ of water, adding

ammonia solution until neutral, and then lcm^6M-acetic acid. ’ •

459

XI. 37 QUANTITATIVE INORGANIC ANALYSIS

B. Dekmiination of chromium as lead chromate (precipitation from hora-
oKcneou-s solution). Disaismit. Use is made of the homogeneous generation
ofchromaie ion produced by the slow oxidation ofchrorniuinflll) by broinatcat
90-95 C in the presence of excess of lead nitrate solution and an acetate buffer.
The crystals of lead chromate produced arc rclatisely large and easily filtered; the
volume of the precipitate is about luilf tliat produced by the standard method of
precipitation.

2cy^ +BrOr t 5H:0 - 2CrO^' - f Hr’ + lOir
BrOr T5Br- +611’ 3Br, (-.3H/)

Pb-^fCrOr -BbCrO.

Cations forming insoluble ciiromates, such as those of silver, barium, mcrcur>(l),
mercury! 11), and bismuth, do not imcrfcrc hcc.nise the acidity is sufficiently high
to prevent tlieir precipitation. Bromide ion from the generation may be expected
to form insoluble silver bromide, and so it is preferable to vep.irale silver prior to
the precipitation. Ammonium salts interfere, owing to competitive oxidation by
broniale. and should be removed by treatmem with sodium hydroxide.

BrOj i-2Kll^* -r. Br -rN, f2fr i alKO

Proccilurc. Use a satnpic solution containing about 50 mg of
cbromium(lll). Neutralise the solution by the .iddition of sodium hydroxide
solution until a preeipitale just begins to form, .Add lOciif’ acetate buffer
solution (6.1/ in aceuc acid and O.o.t/ in sodium aectatc), lOcm^ lead nitrate
solution (3.5 g per lOOcm’). and lUcni’ potassium bromate solution (2.0g per
lOlIcni*). Heal to9(K 95 C: .lUcr generation (of chromate) and precipitation arc
complete (about 45 minutes), as shown by the clear supernatant liquid, cool, filter
on a weighed sintcfed gla.ss or porcelain filtering crueible. wash vviih a little 0.1
percent nitric acid, and dry to e.mstant vvcighi at 120 C. Weigh as PbCrOj.

XI, 37. COBALT. Cobalt may be separated in one of the following forms:

A. Deterniinalinn of cobalt with l-iiiiroso*2-naphthol. DiiaiMhm. 1-
Nitroso-2-naphihol gives a red precipitate witli solutions isfcobalt salts. There is
some doubt as to tiic exact composition of the precipitate; the formula
Co(C,ui!(,OjNlj has been assigned tv> it. but it is probably nut pure. The
complex i-s best eunverted into cobalt sulphate. CoSO^. or into metallic cobalt,
and weighed in either of Ihcsc forms. Eor the limitations of the method, see
Section XI, IIF,

It has been stated that a precipitate of dclmiie composition CofCmUoONOlj
i,s obtained if the cobalt(ll)( 1-30 mg) is lir.si oxidi.sed tocoh.dt(Ul) with a little 30
per ecm liydrogen pertvxide in faintly .icid solution. .Sodium hydroxide solution
(23/) is added until black cobalt(ni) hydroxide conimcnce.s to precipitate; the
latter is dissolved in warm acetic acid and the solution diluted to 200cm-’. The 1-
nitroso-2-naphthol reagent (U)- 20cm^) is added with stirring to the wanii
solution, which i.s then hc;itcd with vigorous stirring until the precipitate
coagulates, 'flic colouretl precipitate is filtcrcii through a weighed poicclain or
sintered glass filtering crucible, washed with a little dnuic acetic acid (H^) and
thrice with hot water. U is dried to constant weiulu at 130 C and weighed as
Co(C,oH,,ONO)j.

An important application is to the .separation ofnickel and cobalt; a double
precipitation is desirable when nickel is present in large amount.

460

GRAVIMETRY XI, 37

Procedure. Dilute the solution containing not more than 0. 1 g of Co as
chloride or as sulphate (1) to 200 cm^, add sufficient concentrated hydrochloric
acid to give a total of 5 cm^ of the concentrated acid in the solution, and warm to
about 80 °C. Add the freshly prepared l-nitroso-2-naphthol reagent (for
preparation, see Section XI, IIF) until precipitation is considered complete:
about 0.25 g of l-nitroso-2-naphthol is required for each 0.01 g of Co. Heat to
gentle boiling with stirring until the precipitate coagulates or settles out. The
supernatant liquid should be clear and yellow. Test whether precipitation is
complete by adding a little more of the reagent to the clear solution. Allow to
stand for 2-3 hours and decant the clear solution through a quantitative filter
paper (e.g., Whatman, No. 541). Wash the precipitate by decantation with a little
hot {ca. 80 °C) dilute hydrochloric acid (1 :2), finally transfer the precipitate to
the filter paper, and wash with hot water until free from acid. Dry the bulky
precipitate at 100-1 10 °C for 1 hour; it will shrink considerably. Place the dried
filter paper and precipitate in a silica crucible, just cover the precipitate with A.R.
oxalic acid (this will prevent sudden decomposition of the complex and
consequent mechanical loss during the ignition process), and heat gently until all
the organic matter has burned off. Ignite the precipitate for a few minutes, allow
to cool, treat with a few drops of concentrated nitric acid to oxidise any residual
carbon and to convert the oxide into nitrate: heat carefully until the excess of
nitric acid has been expelled. Finally, add enough sulphuric acid to convert the
nitrate into sulphate, heat cautiously until the excess of acid has been expelled,
and then for a few moments to incipient redness (450-500 °C). Allow to cool,
moisten with a drop or two of water, and again heat cautiously as before to expel
any free sulphuric acid. Allow to cool in a desiccator and weigh as the sulphate,
C0SO4.

In an alternative procedure, the cobalt is weighed as the metal. Transfer the
dried precipitate and filter paper to a crucible. Ignite the precipitate and filter
paper in the presence of oxalic acid as before: it is important to heat very slowly
at first and subsequently to oxidise any residual carbon with a little concentrated
nitric acid. Fit a crucible-cover on to the crucible, and continue the ignition
in a stream of pure hydrogen for at least 30 minutes. Withdraw the burner
beneath the crucible and, after the crucible is almost at room temperature,
momentarily stop the stream of hydrogen so as to extinguish the flame burning at
the cover. Continue the passage of the stream of hydrogen until the crucible is at
room temperature, and weigh as metallic cobalt. Repeat the treatment with
hydrogen (heating period of 30 minutes in hydrogen) until constant weight is
attained.

Note. 1. A suitable solution for practice in this determination may be
prepared from 0.1 g, accurately weighed, of A.R. cobalt(II) sulphate (clear,
uneffloresced crystals) or pure ammonium cobalt sulphate.

B. Determination of cobalt as cobalt tetrathiocyanatomercurate(II)
(mercurithiocyanate). Discussion. This method is based upon the fact that
cobalt(II) in almost neutral solution forms a blue complex salt Co[Hg(SCN)4]
with a reagent prepared by dissolving 1 mol of mercury(II) chloride and 4 rhols of
ammonium thiocyanate in water. The precipitate is sparingly soluble in water,
soluble in acids and in a large excess of the reagent, soluble in diethyl ether,
chloroform, and carbon tetrachloride, and sparingly soluble in absolute ethanol!
It may be dried at 100-1 10 °C. The following elements interfere: copper
cadmium, zinc, iron(II), iron(III), nickel, manganese(II), bismuth, silver and

461

XI, 38 QUANTITATIVE INORGANIC ANALYSIS

niercur>{ll); iron(lll) may be rendered innocuous by the addition of phosphate.

Proceilurc. The almost neutral sample solution may conveniently contain
35-40 mg of cobalt in 25cm-’ (1) and Ik free from the interfering elements
mentioned above. Add, with constant stirring, 4.3 cm^ of the mercur>(ll)
chloride soltition (2) followed by 5.2cm^ of the ammoniurn thiocyanate reagent
(2). Do not .scratch the sides of the beaker with the .stirring rod. A dark blue
precipitate forms after stirring for I --3 minutes; continue the stirring for a further
2-3 minutes and allow to stand for 2 hours at room temperature. Collect the
pax-ipitate in a weighed .sintcreri glass (poro.sity No. 4) or porcelain filtering
crucible; u.sc the filtrate to assist the transfer of any residual precipitate in the
beaker. Wash the precipitate with 2-3 enf' ofa dilute solution of the pra-ipilating
reai’eni (3) and iinally with 5cin^ icc'Coid water. Dry at 100 C. Weigh as
CoFUgiSCNlJ.

iNok-s. 1, ,A suitable solution for pniclicc analvsis may be prepared by
di.ssoIving about 5.0 g, accurately weighed, pure ammonium coh.dt sulphate in
water and diluting to 50()cin’ in a graduated llask. Lise 25.0cm^ for each
determination.

The cobalt coiucut m.ry Lk rapidlv checked by titration w iih .standard EDTA
svslution m the presence of Xjlenol Orange as indicator (sec Section X, 6t).

2. Sol-Mioit (i); dissolve 5,4 g finely powdered A.R, ractcutifl!) chloride in
IflOcm-’ of distilled water; slight warming may be ncxxssary.

XWm/un (ii); kii;>sol\c b.Og A.K. ammoniurn thiocyanate in 100 cm-* of distilled
water.

It i.s preferable to add tire solutions separ.tlely to the cobalt solution: a slight
e.xcc-s,s (up to It) per cent) of Solution (li) is not harmful. About l.Octn’ of each
solution is required for the pt ecipttation of 10 mg of cobalt. The e.xccsS of the
reagent should not be more than 10- 15 per cent owing to the solubility of the
precipitate in the ammonium mercuriihiocyan.iie .solution.

3. The washing solution rs prepared by arlding i.Ocm^ eacii of Solutions (i)
and (ii) to lOOctn’ of water.

XI, 33. COl'ElvR. Copper may be determined in the following forms;

A. netcrminatioii of copper as coppcr(l) thiocyanate. Discusnun. Thesis
an excellent mciliod, since most thiocyanates of other metals are soluble.
Separation m.iy thus beclfectckl from bismuth, cadmium, arsenic, antimony, tin,
iron, nickel, cobalt, manganese, and zinc. The addition or2- 3 g of tartaric acid is
desirable for the prevention of hydrolysis when bi.smuih. antimony, or tin arc
present. E.xcessivcamounl.s of ammonium salts or of the thiocyanate precipitant
should be absent, as .should also ti.xidising agents; the .solution should only be
slightly acidic, since the .solubility oftlic prccipit.iic increases with da'rcasingpH.
Lead, inea'ury, the precious metals, selenium, and tellurium interfere and
contaminate the precipitate.

The essential experimental condition.s arc;

!. Slight acidity oi the solution with icspcct to hydrochloric acid or sulphuric
acid, since the solubility of the precipitate increases appreciably with
decreasing pH.

2. file presence of a reducing agent, such as sulphurous acid or ammonium
hydrogcnsulphitc, to reduce coppcr(ll) tocoppcr{I).

3. A slight c.xcess of ammonium thiocyanate, since a large e.xccss increases the

462

GRAVIMETRY XI, 38

solubility of the copper(I) thiocyanate due to the formation of a complex
thiocyanate ion.

4 . The absence of oxidising agents.

The reaction may be represented as ; ’■

2Cu2+ THSOj- +H2O = 2Cu+ +HSO4- +2H+

CU++SCN- = CuSCN

The precipitate is curdy (compare silver chloride) and is readily coagulated by
boiling. It, is washed with dilute ammonium thiocyanate solution: a little
sulphurous acid or ammonium hydrogensulphite is added to the wash solution to
prevent any oxidation of the copper(I) salt.

Procedure. Weigh out accurately about 0.4 g of the copper salt (1) into a
250-cm^ beaker, and dissolve it in 50 cm^ of water. Add a few drops of dilute
hydrochloric acid, and then a slight excess (about 20-30 cm^ are required) of
freshly prepared saturated sulphurous acid solution. Alternatively, add 25 cm^
ammonium hydrogensulphite solution : the latter is prepared by diluting to ten
times its volmne the commercial concentrated solution, which has a specific
gravity of 1.33 and contains about 54 per cent sulphur dioxide. Dilute the cold
liquid to 150-200 cm^, heat nearly to boiling, and add freshly prepared 10 per
cent ammonium thiocyanate solution, slowly and with constant stirring, from a
burette until present in slight excess. The precipitate of copper(I) thiocyanate
should be white ; the mother liquor should be colourless and smell of sulphur
dioxide. Allow to stand for two hours, but preferably overnight. Filter through a
weighed filtering crucible (Gooch, sintered glass, or porcelain), and wash the
precipitate ten to fifteen times with a cold solution prepared by adding to every
100 cm^ of water 1 cm^ of a 10 per cent solution of ammonium thiocyanate and
5-6 drops of saturated sulphurous acid solution, and finally several times with 20
per cent ethanol to remove ammonium thiocyanate (2). Dry the precipitate to
constant weight at 1 10-120 °C (3). Weigh as GuSCN.

Notes. 1. A.R. copper sulphate pentahydrate is suitable for- practice in
this determination. 0.4 gram of this contains about 0. 1 g of Cu.

2. Alternatively, but less desirably, the precipitate may be washed with cold
water until the ffltrate gives only a slight reddish coloration ' with iron(III)
chloride, and finally with 20 per cent ethanol.

3. The precipitate, collected in a sintered glass (porosity No. 4) or porcelain
filtering crucible, may be weighed more rapidly as follows. Wash the copper(I)
thiocyanate five or six times with ethanol, followed by a similar treatment with
small volumes of anhydrous diethyl ether, then suck the precipitate dry at the
pump for 10 minutes, wipe the outside of the crucible with a clean linen cloth and
leave it in a vacuum desiccator for 10 minutes. Weigh as CuSCN.

B. Determination of copper >vith benzoin-a-oxime. Discussion. Benzoin-
a-oxime (cupron) is a specific reagent for the determination of copper in
ammoniacal solutions (compare Section XI, IID). A green, heavy, and readily
filterable precipitate is obtained: this is insoluble in water, dilute ammonia
solution, acetic acid, tartaric acid, and ethanol, is slightly soluble in concentrated
ammonia solution, and readily soluble in mineral acids; Precipitation is
quantitative in ammoniacal tartrate solutions : separation can thus be effected
from iron and other metals whose hydroxides are not precipitated in tartrate
solutions. Separation can also be made from cadmium, zinc, cobalt, and nickel,
which are not precipitated in ammoniacal solutions.

463

Xl,39 QUANTITATIVE INORGANIC ANALYSTS

Prinriluri'. Treat the neutral solution, which should be free from am-
monium salts and contain mn mote than 0,05 g copper, with dilute ammonia
solution until a dear blue solution is obtained. Heat to boiling and precipitate the
copper by the addition. dropwKe, of a 2 per cent etlianolic solution of the
reauent. Precipitation is complete when the blue colour of the solution
disappears. Filter the heavy preen precipitate on a weighed sintered glass or
porcelain iilterine crucible, wash w ith hot dilute afnmoni.t solution ( ! 1 lOOj, then
with hot water, and finally with warm ethanol. Dry to constant weight at
105- 1 1 5 C. Wcigli as CufC , , OjN). U is recommended that completeness of
washing be tested fiw by w.i.shine the dry preeipitate again with warm ethanol
followed by hot water. Dry again to eoiisiant '.seight at 105-1 10 C.

XI, 3‘>. GOLD. Detenuination of gold as the metal. iJtscii’iUfJii. Gold is
nearly always determinevi as tiie metal. TTse reducing agents generally employed
are sulphur dioxide, oxalic acid, and ironf 11) sulphate. If nitric acid is present it
must be remosed by repeated evaporation with concentrated hydrochlorie acid,
and the solniion diluted with water. With .sulphurous acid, small amounts of the
plaimum inelaK tpartieui.wly platinum) may be e.irtied down with the pre-
cipitate. It is therefore usually neecssuty to re-x!issol\e the solid in dilute aqua
regia and to re-prccipitaie the gold ; oxalic .icid gives a Iwttcr separation from the
platinum melahs m tiic second precipitation, although the precipilufc is
somewhat finely divided. Iron(ll) sulphate gives s.iji,sraclory results for gold
alone, but diiiicuities arc introduced if the platinum mciai.s arc subsequently to be
determined. Oxalic acid is slow in its action, and yield.s .i precipitate which is
ditlkult to filter.

The best results aie obtained with qumoi as the reducing agent, Precipitation
in hot l,2,)/-h>drochloric acid solution is rapid, liie gold is readily filtered, and
occlusion of the platinum metals is negligible. Precipitation in the cold is
complete in 2 hours. Palladium in tiie lilirate can he precipitated directly with
dimetiiylgly oxime, whilst platinum m the liiir.iie may be determined cither by
evaporating to dryness in order to destroy organic inaUcr and then digesting with
a Imie aqua regia or by icdnciion with sodium formate and formic acid.

Gold may also be separated from hydrochloric acid solutions of the platinum
metals by extraction with diethyl ether or with ethyl acetate (compare Chapter
VT); except m special cases these methods do nut od’er any s[>ccial advantages
over the reduction to the metal.

Procedure The solution should contain not more than 5cm^ con-
centrated hydriKliloric acid per lUOem-* of solution, not more than 0,5-1 gAu,
and be lice from lead, selenium, tellurium, and ilic .dkaline earths. ,\dd 25 cm^ of
u ireslily prepated saturated sulphur dioxide .solution, and digest for 1 hour on
the steam bath in order to coagulate the precipitate. .Add .s-Tocm^ more of the
suljihur dio.xidc solution, and allow to cool. If the cold solution smells strongly ol
sulphur dio.xide, tlic precipitation of gold is complete. .Some of the metal is finely
divided, and it is therefore advisable to make use of a Whatman accelerator or
asliless tablet. Pour the supernatant liquid ihrongli a Whatman No, 42 or 242
filter paper, preferably containing some filter-paper pulp, and transfer as little as
pos.sible ol the precipiiatc to the paper unle.ss one precipitation is thought
sufficient; this will only be the ca-Se if very small amounts of platinum or
palladium are present. Wash well by decantation with hot dilute hydrociiloric
acid ( 1 : 99). Transfer the filler to the beaker, and re-dissolvc the gold in dilute

464

GRAVIMETRY XI, 40

aqua regia; use 8cm^ of concentrated hydrochloric acid, 2cm^ of concentrated
nitric acid, and lOcm^ of water for each gram or less of gold. Filter from the
paper pulp, and wash thoroughly with hot dilute hydrochloric acid (i:99).
Evaporate the filtrate to dryness on the water bath, add 2-3 cm^ of concentrated
hydrochloric acid, and evaporate to dryness again; repeat this operation twice in
order to eliminate all the nitric acid. Treat the residue with 3 cm^ of concentrated
hydrochloric acid, 5 drops of concentrated sulphuric acid, and 75 cm^ of water,
disregard the small amount of gold which may separate, add 25cm^ of a
saturated solution of oxalic acid, and boil for a minute or two. If no further
visible precipitation of gold occurs, digest the solution on the water bath for at
least 4 hours. Filter off the gold through a filter paper (as described above), and
wipe the inside of the beaker with small pieces of quantitative filter paper to
ensure that all the metal is transferred from the beaker; wash well with 1199
hydrochloric acid. Transfer the filter to a weighed porcelain or silica crucible,
bum off the paper carefully, and ignite to constant weight. Weigh as Au.

Procedure B. The solution must be free from nitric acid, be about 1.2M
with respect to hydrochloric acid (ca. 5 cm^ of concentrated hydrochloric acid in
50 cm^ of water), and contain up to 0.2 g of Au in 50 cm^. Heat the solution to
boiling, add excess of 5 per cent aqueous quinol solution (3 cm^ for every 25 mg
of Au), and boil for 20 minutes. Allow to cool, and filter either through a weighed
porcelain filtering crucible or through a Whatman No. 42 or 542 filter paper;
wash thoroughly with hot water. The small particles of gold remaining in the
bottom of the beaker (easily seen with a small flash lamp) are best removed with
pieces of ashless filter paper. Ignite the porcelain filtering crucible to constant
weight. If filter paper is used, transfer to a weighed porcelain or silica crucible,
and complete the determination as described in Procedure A.

XI, 40. IRON. Iron may be determined in the following forms :

A. Determination of iron as iron(III) oxide by initial formation of basic
iron(III) formate. Discussion. The precipitation of iron as iron(III) hydroxide
by ammonia solution, etc., and its conversion into iron(III) oxide, in which form
it is weighed, is fully described in Section XI, 21. Precipitation with ammonia
solution yields a gelatinous precipitate which is somewhat difficult to wash and to
filter; this difficulty is largely overcome by the addition of a few drops of
hydrazine hydrate after the ammonia solution.

There are, however, three methods of precipitation which yield iron(III)
hydroxide in a relatively dense and granular form, which is easily washed and
filtered. These include the precipitation of iron(III) as basic iron(III) formate
from homogeneous solution by hydrolysis of urea in a hydrochloric acid solution
containing formic acid. The precipitate thus obtained is denser, more readily
filtered and washed, and adsorbs fewer impurities than the precipitate obtained
by other hydrolytic procedures. Ignition yields iron(III) oxide. The pH at which
basic iron(III) formate begins to precipitate depends upon several factors, which
include the initial iron and chloride concentration: a high concentration of
ammonium chloride is essential to prevent colloid formation. It is important to
use an optimum initial pH to avoid a large excess of free acid, which would have
to be neutralised by urea hydrolysis, and yet there must be present sufficient acid
to prevent the formation of a gelatinous precipitate prior to boiling the solution :
ideally, a turbidity should appear about 5-10 minutes after the solution has
begun to boil. For iron contents of 5 mg to 55 mg per 100 cm^, the optimum

465

XI,40 QUANTITATIVO INORGANIC ANALYSIS

initial pH is between 2.00 to L70, Some reductioti occurs during the
precipitation (due to the presence of both formate and chloride): re-oxidation of
ironfil) to iron(ill) ts easily ellccted by tlic addition oi hjdrogcn peroxide
towards the end of ilte procedure. Precipitation a,s basic fonnale enables iron to
be separated from inunganesefn), cobalt, nickel, copper, zinc, cadmium,
niagne-siurii. calcium, and barium. Wlicn copper is present the solution must be
eoolcd hetbre hydrogen peroxide is added, otherwise the vigorous decom-
position of the liydrogen pero-xidc may result in loss of some of the solution.

Attention is directed to the fact iliat if ignition is carried out in a platinum
crucible at a temperature above 1 K.W C some reduction to the oxide FcjO^ may
occur, and at teinperatuies .ibuve I20U C .some of the oxide may be reduced to
the metal and alloy with the plutinmn. This accounts in part for the con-
tamination ofllicplatinumcnicibleby iron which sometimes rx'curs in analytical
work. No magnetic oxide of iron is produced if silica crucibles are employed for
the ignitions.

Putccilurc. I'or practice in this determination A.R. ammonium ironflll)
sulphate (iron alum) may be used. Weigh out accurately about l.Og A.R. iron
alum, dissolve it in dilute hydrochloric acid, add 2.0cm^ formic acid (sp. gr. 1.30:
CO. ‘JO percent), 10 g ammonium chloride, and -1.5 g mea (as a 10 percent aqueous
solution); mix well. Dilute to about ,»50cm*. Adjust the pH to 1.80 (use a pH
meter) by the addition of hydrochloric acid.* Dilute to 400 cm ', in.sert a boiling
rod, and boil gently for about ‘J>) minutes or until a pH of about 4.0 is reached.
Add 5 cm^ of3 [sercent liydrogcnpcroxidcsolulion and boil fora further 5ininuies,
Then add 10 cm'* 0.02 iser cent gel.itm sohition: the latter will improve the
filtering and wa.shing properties of the precipitate, l-'ilter on a Whatman No. 40
or No. 540 filter paper and wash the prccipil.ite fifteen times with hot 1 perectu
ammonium nitrate solution .uljusled to pH 4. Kcinovc as much as possibleof the
precipitate ,tdhcrmg to the walls of the beaker with the aid of a rubber-tipped
stirring rod.

Di.ssolve any film adhering tenaciously to the w,ills of the beaker by adding
-1-5 cm' of concent rated hydrochloric acid, cover with a clock glass, andrciiu.x
gently for a few minutes. Then w.ish the beaker, chick glass, and stirringrod with
25 cm' distilled water, add a few drops of methyl ted md!C.Uor and then dilute
ammonia solution dropwi.se until the colour of ilic solution is a di.slinct yellow.
Boil for 2 3 minutes to coagulate the precipitate, filter, and wash on a separate
filler paper, Blaee both filler p.ipers m a weighed porcelain or .silica crucible, char
the filter papers over a small tlamc. and then ignite at a red heat or in an claaric
miilile furnace at 85U C. Healing for ! hour is usually .sutiicicni. Weigh as Fe^Oj.

B. Delcrniinalioii of iron with cupferron uii J subsequent weighing as iron(Ul)
o.xtde. Dm uAsioii. Cupferron, the ammonium salt of nitrusophenylhydro.xy-
laniine, C<,H 5 N(NO)'ONH 4 , precipitates iron, tin, uranium(lV), vanadium,
titanium, and /irconium from strongly acid sohitioas. thus alTording a sep-
aration fiom aluniiniuin. chromium, hcryllimn, phosphoru.s, boron, manganese,
zinc, nickel, cobalt, and hexavalent uranium (compare Section .Kl, UB). Copper

• Allcriuuscty, aUd puie .uiucous ammonia soluium (1 ■ 1) by me.ir>s of a dropper pipeue until a
dcfmiicpiccipii,itc()f ironttll) hjdfoodejiwl bcpiiu lo form. Sow .uidconccntniicdhydroetitoric
acid drops', ISC, stitrinj; and allunliij; tosiand fora mmuie or Isso after each 2- J drops, uniilacicar
solution i> obtained ; ihcn add 1 .5 cin^ of eonecniralcd li> droelilorie acid and mix s'cH.

466

GRAVIMETRY XI, 41

and thorium must be precipitated from weakly acid solutions. Several metals
(e.g., lead, silver, mercury, bismuth, tungsten, and cerium) interfere, but most of
these may usually be removed by other methods— for example, by hydrogen
sulphide in acid solution. The precipitate cannot be weighed as such, but must be
ignited to the oxide. The ignition must be very carefully carried out in the early
stages in order to avoid mechanical losses, for wet precipitates tend to liquefy and
effervesce, whilst, dry precipitates give off considerable volatile matter. The
precipitate is rather bulky, and the amount of material taken should therefore be
such as to yield 0. 1-0.2 g of oxide.

Procedure. The solution ( 1) ( 1 50-200 cm^) should contain about 0. 1 g of
Fe in the iron(III) state and be strongly acid with hydrochloric acid or sulphuric
acid. To the cooled solution {ca. 10 °C) add a freshly prepared, filtered 5 per cent
aqueous cupferron solution (2) slowly and with, constant and vigorous stirring
until no further formation of a brown precipitate takes place. The formation of a
white precipitate of nitrosophenylhydroxylamine indicates when the reagent is
present in excess. Do not warm, since the reagent is rapidly decomposed in hot
acid solution. Add a Whatman accelerator (or a third of a Whatman ashless
tablet), stir for 2-3 minutes, and without further delay filter through a Whatman
No. 41 or 541 filter paper, preferably supported upon a Whatman filter cone.
Wash several times with 10 per cent by volume of hydrochloric acid containing
1 .5 g of cupferron dm" then twice with 5Af-anunonia solution to remove excess
of cupferron, and finally once with water. Ignite the precipitate with the paper in
a weighed porcelain, silica, or platinum crucible, very gently at first until all the
organic matter is destroyed, and then strongly to constant weight. Weigh as

FejOj.

Notes. 1. A suitable solution fpr practice may be prepared by weighing
out accurately 0.8-0.9g A.R. iron alum, dissolving it in ISOcm^ water, and
adding 25 cm^ concentrated hydrochloric acid.

2. Only freshly prepared solutions should be employed, because the solution
only keeps for a day or two. The dry reagent should be kept in a cool, dark place,
and preferably with a bag of ammonium carbonate suspended in the bottle.

XI, 41. LEAD. Lead may be determined in a number of forms in addition to
that of lead chromate already described (Section XI, 22). Two useful procedures
follow:

A. Determination of lead as molybdate. Discussion. This is an excellent
method, since the substance has a high molecular weight, is less soluble than the
sulphate, and suffers no change upon ignition. Substances which form insoluble
molybdates (e.g., the alkaline-earth metals, copper, and cadmium), which are
easily hydrolysed (e.g., tin or titanium), and which form insoluble compounds
With lead (e.g., chromates, arsenates, or phosphates) must be absent.

Procedure. Weigh out accurately about 0.30 g of the lead salt, dissolve it in
200 cm^ water, and add 4 drops concentrated nitric acid. Heat to boiling, and
slowly add from a burette or pipette, with stirring, a 2.5 per cent aqueous.solution
of ammonium molybdate. When precipitation appears to be complete, boil
for 1 minute, allow the precipitate to settle, and add a few drops of the precipitant
to the supernatant liquid. If a precipitate forms, repeat the process until the
molybdate is present in slight excess. When precipitation is complete,
add dilute ammonia solution (112) dropwise until the solution is neutral or
Slightly alkaline to litmus or methyl red. Acidify with a few drops of acetic acid,'

467

XI, 42 QUANTITATIVE INORGANIC ANALYSIS

and allow lo stand for a few minutes. Decant the .supernatant liquid through a
weighed porcelain or silica filtering crucible, and wash tlie precipitate three or
four times by decantation with 75-cni^ portioas of 2 percent ammonium nitrate
.solution. Traii-sfer the precipitate to the filter, and svash until the washings give
no test for molybdenum (c.g., no brown precipitate with potassium

hexacyanoferratel II ) solution). Place the filtering crucible inside a nickel crucible

or upon a crucible-ignition dish, and gradually heat to dull redness. Maintain the
crucible at dull redness for 10 minutes, cool in a desiccator, and weigh. Repeat the
heating, etc., until constant weight i.s attained. Alternatively, heat in an electric
muITle furnace at 500-600 "C to coirstant weight. Weigii as PbMoO^.

B. Determination of lead as thesalicylaldo.vimate. Discussion. Lead may
be precipitated in strongly ammoniacal solution (pH 9.3 or higher) with
salicylaldchydeo.xime as the lead coinple.v Pb{C-,HjOjN); if should therefore be
po.ssiblc to separate lead from silver, cadmium and zinc, the saiicylaldo.ximate of
which arcsoluble in ammoniacal solution (see Section XI, HE).

Proccilurc. To a .solution (at. 25 cm') of lead nitrate or lead acetate
containing about O.I g of Pb, add lOcm^ freshly prepared I |3er cent salicylal-
dehydco.xime solution (for preparation, see Section .\I, 1 1 E), dilute to 50cm' .and
add 12.5 cm' concentrated ammonia solution. Stir the resulting precipitate for 1
hour and allow to settle. Decant the supernatant licjuid through a sintered glass
crucible (porosity No, 4). wash the precipitate by decantation until free from
.saiicylaldcliydeoxirne(as shown by the absence of a colour with iron(ni) chloride
solution), dry at 105 C for t hour, and weigh us Pl)(C.HjO,N).

XI, 42. LITTllU.M. It is usually nccc.ssary to determine lithium in the
presence of sodium and/or potassiumv The following procedures are suitable for
this:

A. Ueterminatiun of lithium in the presence of sodium and potassium by
c.xtraetiorr wiflt orgattic solvenfs. Discussion. This procedure is dependent
upon the fact that lithium chloride is very soluble in many organic solvents,
whilst the chlorides of sodium and potassium are only very slightly soluble. A
number of solvents have been suggc.sied as being .suitable for this purpose,
including dioxan, hexanol, 3-irwibyl-l -butanol and 2-ciliyl he.xanol.The foilovv-
ing fable gives the solubilities expressed as grams di.ssoivcd by lOOcin' of the
anhydrous solvent at 25 C.

.VnieUijI-l-tHilaiiol

Ilrxaiiul

I-iQ

7.a

5.8

.v.o

N.iCl

oooifi

O.OOOS

0 001)1

KCI

O.IHKKi

O.lXXIlW

<0.00001

The prcferretl solvent is 2-cthylhexanol.

Procedure. Treat a concentrated solution prepared from 0.3-0.4g or less
of the mi.xed chlorides, accurately weighed, with a suitable volume of 2-
cthylhc.xanol, introduce a little platinum foil or a few fragments of porous
porcelain lo prevent bumping, and distil until the water has passed over and the
boiling point becomes constant (175-180 C) for sonic time. Sodium and
potassium chlorides are deposited, and lithium chloride is dehydrated and held in
solution. Allow to cool, filter through a sintered gla.vs filtering crucible, and wash

468

GRAVIMETRY XI, 43

thoroughly with successive small volumes of the anhydrous alcohol. Dry the
crucible at 200-210 °C to volatilise the residual, solvent, and weigh. The loss in
weight is due to the lithium chloride.

If. the weight of the lithium chloride exceeds 20 mg, a second extraction is
necessary in order to remove the small quantity of lithium hydroxide present in
the residual solid (formed by hydrolysis, at the boiling point of the .2-
ethylhexanol) : the solid must be dissolved in a little water containing a few drops
of hydrochloric acid. .

B. Determination of lithium as lithium aluminate. Discussion. Lithium
may be determined as litluum aluminate by precipitation with excess of sodium
aluminate solution in the cold, the final pH of the solution being adjusted to
12.6-13.0. The precipitate is washed with water until free from alkali and weighed
as 2Li20-5Al2O3 after heating at 500-550 °C. The solubility in water is 0.008 g
dm“^ at room temperature; it is 0.09 g dm“^ at pH 12.6.

Procedure. The sample solution (20 cm^) may contain up to 10 mg of
lithium, and the pH should be about 3.0. Add 40 cm^ of the cold reagent (1) for
each lOmg of lithium. Adjust the pH to 12.6 by the addition of lAf-sodium
hydroxide solution: use a pH meter. Allow to stand for 30 minutes, and collect
the voluminous precipitate in a porcelain filtering crucible. Wash with small
volumes of ice-cold water until the washings are no longer alkaline to
phenolphthalein. Ignite at 500-550 “C in an electric muffle furnace. Weigh as
2Li20-5Al203.

Note. 1. Prepare the precipitating reagent by dissolving 5.0 g A.R.
aluminium potassium sulphate (potash alum) in 90 cm^ warm water. Cool and
add dropwise with stirring, while cooling in ice, a solution of 2.0 g sodium
hydroxide in 5.0 cm^ water until the initially formed precipitate redissolves. After
standing for 12 hours, filter, adjust the pH to 12.6, and dilute to 100 cm^. with
water.

XI, 43. MAGNESIUM. Methods for the determination of magnesium as
ammonium magnesium phosphate hexahydrate and as magnesium pyrophos-
phate have already been given in Section XI, 23. The following method is also
useful :

Determination of magnesium as the 8-hydroxyquinaldinate. Discussion.
Magnesium may be precipitated by 8-hydroxyquinaIdine (2-methyloxine)
m ammoniacal solution (pH at least 9.3) as the complex Mg(CioH80N)2,
and weighed as such after drying at 130-140 °C. Numerous ions
interfere (see Section XI, IIP).

Procedure. The solution (1 50-200 cm^) may contain up to 0.05 g of Mg.
Add 3cm^ of the 8-hydroxyquinaldine (2-methyl-oxine) reagent (1) for every
0 mg of magnesium present, and then add concentrated ammonia solution until

further precipitate forms. Digest the solution at
-80 C for 20 minutes arid filter through a sintered glass or porcelain filtering
crucible. Wash the precipitate with hot water and dry to constant weight at
130-140 “C. Weigh as Mg(CioH80N)2.

in I reagent is prepared by dissolving 5 g of 8-hydroxyquinaldine

g 01 glacial acetic acid and diluting to 100 cm^ with water.

XI, 41/45 QUANTITATIVE INORGANIC ANALYSIS

XI, 44. MANGANESE. Detcrniination of manganese as the ammonium
phospliate or as the pyrophasphatc. Disaission. The only method which is at
all widely used for the gravimetric estimation of manganese is the precipitation as
ammonium manganese phosphate, in slightly aramoniacal

solution containing excess of ammonium salts. The precipitate may be weighed
in this form after drying at 100 -105 'C, or it may be ignited and subsequently
weighed as manganese pyrophosphate, MujP,0,. The latter procedure is by far
the better one. The method is, however, of limited application because of the
interfering intluencc of numerous other elements. Titrimclric methods are
generally preferred (sec Cliapter X); the potcntiomctric determination of
manganese (see Section .XIV, 28) may also be recommended.

Procedure. The solution (200cin^) should be slightly acid, contain not
more than 0.2 g of Mn in 200cin^, and no other cations except those of the alkali
metals ( 1 ). .Almost neutralise the solution with dilute ammonia solution, add 20 g
of ammonium chloride and a considerable excess of diammonium hydrogen
phosphate(NH 4 ),HPO.i(say, 2goflhe solid). Ifa precipitate forms at this point,
dissolve it by the addition of a few drops of 1:3 hydrochloric acid. Heat the
solution almost to boiling (9U -95 C). and add dilute ammonia solution (1 13)
dropwi.se and witli constant stirring until a precipitate (Mn^fPO^),) begins to
form; immediately suspend the addition of the alkali. Continue the heating and
stirring until the precipitate becomes crystalline (MnNH^PO^.H.O). Then add
another drop or two of ammonia solution, stir as before, etc., and so continue
until no more precipitate is produced and its silky appearance remains
unchanged. The precipitate must be maintained at 90-95 C throughout; a large
excess of ammonia .solution mu.st be avoided. Allow ihesolulion to stand at room
temperature (or, better, at O'C) for 2 hours. Filter through a quantitative liltcr
paper or tlirough a weighed porcelain filtering crucible, and wash the precipitate
with cold, 1 percent ammonium nitrate solution until free from chloride. Do' at a
gentle heat, ignite at as low a temperature as possible until the carbon is oxidised
(2). and then heal at 700-800 C (in an electric crucible furnace or within a larger
nickel crucible) to constant \vcig)ii. Weigh as MnjPjO,. Alternatively, but less
desirably, the precipitate in llie porcelain lillering crucible may be dried at
100-105 '"’C to constant weight and weighed as Mn N H^PO^.HjO ; in this case, a
Gooch or a sintered glass lillering crucible may also be u.scd.

Notes. 1. A suitable solution for practice may be prepared by one of the
following methods:

(u) Dissolve 0.7 g, accurately weighed, A.R. manganeselll) sulphate
MnS 04 , 4 H 30 in 200cm^ of water.

(/)) Dissolve 0.5 g. accurately weighed, A.R. potassium permanganate in very
dilute sulphuric acid, and reduce the solution with sulphur dioxide or with
ethanol. Remove the excess of sulphur dioxide or of acetaldehyde (and
ethanol) by boiling. Dilute to 200 cm-*.

2. These remarks apply, of course, when filter paper is used.

XI, 45. MERCURY. Mercury may be determined in the following forms:

A. Deterniinalion of mercury as sulphide. Discussion. The precipitation
of mercury as mcrcuryd!) sulphide by hydrogen sulphide in hydrochloric acid
solution is an accurate procedure in the absence of copper, cadmium, tin, zinc,
and thallium; the latter metals complicate reactions which arc based upon the
behaviour of pure mercury! 11) sulphide. Unless the experimental conditions

470

GRAVIMETRY XI, 46

detailed below are strictly followed, the precipitate is liable to be contaminated
with a little sulphur, which must be removed by extraction with carbon
disulphide. Oxidising agents (nitric acid, chlorine, iron(III) chloride, etc.) must
be absent. ■ .

Procedure. Weigh out accurately about 0.15g of the mercury(II) salt (1),
dissolve it in 100 cm^ of water, and add a few cm^ of dilute hydrochloric acid.
Saturate the cold solution with washed hydrogen sulphide (2), allow the
precipitate to settle, and filter through a weighed Gooch, sintered glass or
porcelain filtering crucible. Wash the precipitate with cold water (3), and weigh it,
as HgS, after drying at 105-1 10 °C.

Notes. 1 . A.R. mercury(II) chloride is suitable. Alternatively, the solution
should contain not more than 0.1 g of mercury(II) per 100 cm^, and should be
free from oxidising agents.

2. The colour of the mercury(II) sulphide precipitate will become perfectly
black as soon as the liquid is saturated with the gas.

3. If the presence of sulphur is suspected, the precipitate is washed with hot
water, ethanol, carbon disulphide, or ethanol + diethyl ether, and then dried at
105-1 10 °C.

B. Determination of mercury as mercury(II) thionalide. Discussion.
Thionalide CioH^NHCOCHj-SH may be used for the quantitative pre-
cipitation of mercury(II) as Hg(Cj 2 HioONS) 2 . Sulphate does not interfere.
Attention is drawn to the following experimental points :

1. The chloride ion concentration of the solution should not exceed 0.1 Af; the
results are high if the chloride ion concentration is excessive.

2. If nitric acid solutions of mercury(II) nitrate are used, the latter must be
converted into mercury(II) chloride by the addition of at least an equivalent
amount of chloride ion.

3. A three-fold excess of reagent should be employed.

Procedure. The sample solution may contain 5 to 75 mg of mercury(II). A
solution, prepared from A.R. mercury(II) chloride and containing, say, 20 mg
mercury in 1 50 cm^ of water, may be used for practice in this determination. Heat
the solution to 80-85 °C and add, with constant stirring, a three-fold excess of a 1
per cent solution of thionalide in acetic acid. The precipitate coagulates upon
stirring. Filter the hot solution through a sintered glass filtering crucible
(porosity No. 3) which has been preheated by pouring hot water through it. (The
use of a warm filtering crucible is essential ; the separation of thionalide in the
pores of the sintered plate of the crucible, which would render filtration difficult,
is thus avoided.) Wash with hot water until free from acid, and dry to constant
weight at 105 °C. Weigh as Hg(Ci 2 Hi(,ONS) 2 .

46. MOLYBDENUM. Molybdenum may be determined in the following
forms •

XI, 46 QUANTITATIVE INORGANIC ANALYSIS

iron, chromium, vanadium, tungsicn, silicon, phosphorus, arsenic, antimony, tin,
and titanium interfere.

Procedure. Weigli out accurately about 0.4 g ferro-molybdenum (1),
dissolve it in lOcm^ concentrated hydrochloric acid and 2cm^ concentrated
nitric acid, evaporate to 2-3 env*, dilute to 50 cm^, and transfer to a separatory
funnel. Dissolve 5 g sodium liydro.xidc in 20()cm^ water in the original beaker,
heat to boiling, and run in the solution from the separatory funnel dropwiseand
with constant stirring. Rinse out tlie funnel twice with boiling water, and add the
washings to tlie main solution. Filler olT the precipitated iron(lll) hydro.xide and
wash with hot water. Dissolve the precipitate in the minimum volume of dilute
hydrochloric acid, and re-prccipitate by slowly pouring into a solution of about
4 g sodium hydroxide in lOOcm^ of water. Filler oITthe precipitate. wa.sh with hot
water, and add the filtrate and washings to the main .solution. Acidify with acetic
acid, add SOcnF ofa 50 percent solution of ammonium acetate, and make up to
500 cm-* in a graduated flask. Remove 250cm’ oflhe solution (2), heal to boiling,
and maintain near the boiling point with a small flame: add from a burette a
solution of lead acetate (containing 4 g of the salt and 1 cm’ of glacial acetic acid
per 100cm’) dropwise and with constant stirring. When a slight excess of the
precipitant has been added, the milky solution clears appreciably. When this
occurs, boil for 2 • 3 minutes whil.si the solution is stirred, allow to settle, and add
a few drops of the reagent to see if precipitation is complete, A large excess of
precipitant should be avoided. Digest on the steam bath for 15-30 minuie.s.
Decant the clear solution through a weighed Gooch or porcelain filtering
crucible, wash by decantation three or four times with 75-cnr portions of hot 2
per cent ammonium nitrate solution, transfer the precipitate to the filter, and
wash until the soluble salts have been removed. Dry and ignite the precipitate at a
dull red heat {cu. 600 C) ;is described in Section XI, 41 A. Weigh as PbMoO^.

Notes. 1. The Bureau of Analy.scd Samples ‘Ferru-Molybdenum, No.
231’ (a British Chemical Standard) is suitable.

2. If molybdenum is being determined in a simple .salt, e.g., in A.R. molybdic
acid or molybdic anhydride, commence at this point. The solution should
contain about 0.1 g Mo in 200cm’ and may be prepared as follows. Dissolve
0.15g. accurately weighed, A.R, molybdic acid or anhydride, in 50cm’ dilute
ammonia solution, acidify with acetic acid, add 25 cm’ of a 50 per cent solution
of ammonium acetate, and dilute to 200cni’.

B. Determination of molybdeiunn with oxinc. Diseieydon. Molybdates
yield sparingly .soluble orange-yellow molybdyl 'oxinatc' with oxinc solution ; the
pH of the solution should be between the limits 3.3-7.6. The complex differs from
other 'oxinates' m being insoluble in organic solvents and in many concentrated
inorganic acids. The freshly precipitated compound dissohes only in
concentrated sulphuric acid ami in hot .solutions of caustic alkalis. Tliis
determination is of particular interest, as it allows a complete separation of
molybdenum and rhenium.

Procedure. Neutralise the solution of alkali molybdate, containing up to
0.1 g of Mo, to methyl red, and then acidify with a few drops of ,\/-sulphuric
acid. Add 5cm’ 2A/-ammonium acetate, dilute to 50-100cm’, and heal to
boiling. Precipitate the molybdenum by the addition of 3 per cent solution of
oxine in dilute acetic acid (Note 1), until the supernatant liquid becomes
perceptibly yellow. Boil gently and stir for 3 minutes, filter through a filtering
crucible (sintered glass or porcelain), wash with hot water until free from the

472

GRAVIMETRY XI, 47

reagent, and dry to constant weight at 130-1 40 °C. Weigh as Mo02(C9H60N)2.

Notes. 1. The oxine reagent may be prepared by dissolving,4 g A.R. oxine
in 8.5 cm^ of warm A.R. glacial acetic acid, pouring into 80 cm^ water, and
diluting to 100 cm^.

XI, 47. NICKEL. The determination of nickel as nickel dimethylglyoximate
has already been described in Section XI, 24 ; the following methods are also
useful : -

A. Determination of nickel in the presence of copper with salicylaldehyde-
oxime. Discussion. The complex is precipitated in neutral or very faintly acid
solutions (best at pH = 7) in contrast to that of copper, which is formed in the
presence of acetic acid. The experimental details are similar to those for copper,
except that the solution must be neutral or very faintly acid. Iron(III) interferes,
and should therefore be absent.

Procedure. Treat the solution, free from mineral acid and containing both
nickel (not more than 0.03 g) and copper (about 0.03 g), with 1 g sodium acetate
and 10 cm^ glacial acetic acid per 100 cm^ of solution.* Add excess of
salicylaldehydeoxime reagent over the quantity required to precipitate both
metals, and stir, the solution vigorously during the addition. Filter off the
precipitated copper complex on a weighed filtering crucible (Gooch, sintered
glass, or porcelain), wash well with cold water, and dry at 100-105 °C to constant
weight. Weigh as Cu(C7H602N)2. Add dilute ammonia solution to the filtrate
and washings (diluted to 300-350 cm^) until the solution remains very faintly
acidic. Stir thoroughly to coagulate the precipitate of nickel salicylaldoximate.
Filter through a weighed sintered glass or porcelain filtering crucible, wash with
cold water until the washings give no coloration with iron(III) chloride solution,
and dry at 100 °C to constant weight. Weigh as Ni(C7H602N)2.

B. Determination of nickel by the pyridine method. Discussion. This is a
rapid method with similar advantages and limitations to those described for zinc
(Section XI, 62). A determination may be completed in about 30 minutes.

Procedure. The solution (lOOcm^) should contain about 0.1 g nickel (1)
and be neutral in reaction. Stir in 0.5-1.0 g A.R. ammonium thiocyanate, heat to
boiling, add 1-2 cm^ pure pyridine, and immediately remove the flame. Stir for
2-5 seconds until the precipitate commences to separate in sky-blue prisms. (The
precipitate separates immediately or after standing a short time, according to the
quantity of nickel present.) When cold, filter through a weighed sintered glass or
porcelain filtering crucible, and use Solution 1 to assist in the transfer of the
precipitate to the crucible. Wash four to five times with Solution 2, then twice
with 1-cm^ portions of Solution 3, and finally five or six times with i-cm^ portions
of Solution 4. Dry in a vacuum desiccator at room temperature for 10 minutes
and weigh. Repeat the drying until the weight is constant. Weigh as
[Ni(C5H5N)J(SCN)2 (2).

Solution 1. lOOcm^ water containing 0.4 g NH^SCN and 0.6 cm^ pyridine.

Solution 2. 61.5cm^ water, 37.0cm^ of 95 per cent ethanol, 0.1 g NH^S.CN,
and 1.5 cm^ pyridine.

Solution 3. 10 cm^ absolute alcohol and 0.5 cm^ pyridine.

Solution 4. 20 cm^ diethyl ether (sodium dried) and 2 drops pyridine.

The best pH range for the precipitation of copper in the presence of nickel is 2.6-3. 1 .

XF, 48/49 QUANTITATIVE INORGANIC ANALYSIS

Notes. 1. For practice in titis estimation, employ 0.3 g, accurately
weighed, A.R. nickel sulphate or pure ammonium nickel sulphate.

2. For further experimental details, see under Zinc.

XI, 48. PALLADIUM. Palladium may be determined in one of the following
form.s :

A. Determination of palladiunt with dimetliylglyoxime. Discm.sioii. This
is one of the best method.s for the determination of the element. Gold must
be absent, for it precipitates as the metal even from cold solutions. The
platinum metals do not. in general, interfere. Moderate amount,s of platinum
cause little contamination of the precipitate, but with large amounts a second
precipitation is desirable. The precipitate is dccompo.scd by digestion on the
water bath with a little aqua regia, and diluted w iih an equal volume of water; the
resulting solution is largely diluted with water, and the palladium re-precipitated
with dimeihylgiyo.xime.

An objection to the precipitation of palladium with dimeihylglyoxime is the
voluminous character of the precipitate. Hence if much palhidium is present, an
aliquot part of the solution should be used.

Procaliirc. 'flic solution should contain not more than 0. 1 g Pd in 250 cm^,
be 0.253/ with respect to hydrochloric or nitric acid, and be free from nickel and
gold. Add, at room temperature, a 1 per cent solution of dimelliylglyo.xirae in 95
per cent ethanol. Use 2 5 enn* of the reugeni for every 1 0 mg of palladium. Allow
the solution to stand for 1 liour, and then filter llirougli a weighed filtering
crucible (Gooch, sintered glass, or porcelain). Te.st the filtrate with a little of the
reagent to make sure that precipitation is complete. \Va.sh the orange-yellow
precipitate of palladium dimcihylglyo.ximatc thoroughly, fir.st with cold water
and then with hot water. Dry at llO'C to constant weight. Weigh as
Pd(C4HtOjN,>,.

B. Dcterniinalion of palladium with cyclohc.\ane-L2-dlonedio.vimc. Dis~
cuisioit. Cyclohe.\anc-l,2-dionedio,xime (nioxiine) yields a highly insoluble
yellow compound, Pd(C„H.^02Nn),, with palladiiiin saltsat pH values between 1
and 5 (see Section .\I, 1 lA); it can be filtered from the hot solution after a brief
digestion period. The reagent, unlike dimcthylglyoximc, is soluble in water, and
hence the palladium precipitate is unlikely to be contaminated with excess of
reagent. The precipitate i.s rather bulky, so that dctcrniinalions are conducted
with quantities not exceeding 20-30 mg of palladium. Common anions do not
interfere, nor do beryllium, aluminium, lanthanum, uranium(Vl). and the
alkaline-earth ions. Amounts of platinum (up to that of the palladium) do not
interfere; gold{l) at 60 ' C is partially reduced to metallic gold.

Procedure’. The solution (volume about 200cm^) may contiiin 5-30mg
palladium : the pi 1 may vary from 1 to 5. Heat the solution to 60 'C, add slowly
from a graduated pipette with stirring 0.50cm^ of a 0.8 per cent aqueous solution
of nioxiine for each milligram of Pd present. Digest the solution with occasional
stirring for 30 minutes at 60 C, filter through a sintered glass or porcelain
filtering crucible, and wash well with hot water. Dry at 1 10 "'C to constant weight,
and weigh as PdfColL^OiN,),.

XI, 49. PLATINUM. Platinum is determined preferentially as the metal.

Determination as metallic platinum. Disais.'U'on. The platinum solution is
treated with formic acid, best at pH 6, and the precipitated platinum weighed. A

474

GRAVIMETRY XI, 50

2 per cent solution of hypophosphorous acid may also be used as the reducing
agent.

Procedure. In this determination any excess of nitric and/or hydrochloric
acid present must be removed. Evaporate the solution of platinum, containing no
other platinum metals (ruthenium, rhodium, palladium, osmium, and iridium)
or gold, to a syrup on the steam bath so as to remove as much hydrochloric acid
as possible. If nitric acid was present, dissolve the residue in 5 cm^ of water, heat
on the water bath for a few minutes, add 5 cm^ of concentrated hydrochloric
acid, and again evaporate to a syrupy consistency. Dissolve the residue in water,
and dilute so that the solution does not contain more than 0.5 g of Pt in 100 cm^.
For each lOOcm^ of solution, add 3 g of anhydrous sodium acetate and 1 cm^ of
formic acid. Heat on the boiling water bath for several hours. Filter through a
quantitative filter paper. Add a little more sodium acetate and formic acid to the
filtrate and digest in order to ensure complete precipitation. Wash the precipitate
with water until free from chloride, dry and ignite the filter paper in contact with
the precipitate to constant weight. Weigh as metallic Pt.

XI, 50. POTASSIUM. Potassium may be determined in one of the following
forms :

A. Determination of potassium as dipotassium sodium hexanitritoco-
baltate(in) (cobaltinitrite). Discussion. By precipitation of potassium solu-
tions with sodium hexanitritocobaltate(III) reagent under the experimental
conditions given below, a quantitative yield of dipotassium sodium hexani-
tritocobaltate(III) is obtained. This method is applicable in the presence of
sulphate.

Formerly precipitation was made in an acetic acid solution with a reagent
prepared by mixing a solution of cobalt nitrate or acetate in dilute acetic acid and
one of sodium nitrite in water. There is some evidence that the composition of the
precipitate may, upon occasion, vary slightly from the formula given below. If
precipitation is made from a nitric acid solution by a solution of sodium
hexanitritocobaltate(III), Na3[Co(N02)6], the heavy, crystalline precipitate
invariably has the the composition K. 2 Na[Co(N 02 ) 6 ]H 20 for amounts of
potassium from 2 to 15 mg in lOcm^ of solution. Acidification with dilute nitric
acid tends to prevent decomposition of the nitrite. The precipitate may be
weighed as such after drying at 100-1 10 °C for 2 hours, or it may be converted
into potassium perchlorate.

Procedure. Weigh out accurately 0.03-0.04 g A.R. potassium sulphate
and dissolve it in 10 cm^ water. Add 1 cm^ of M-nitric acid and a freshly
prepared solution of 1 g A.R. sodium cobaltinitrite in 5cm^ water, mix, and
allow to stand for 2 hours. Filter through a weighed sintered glass or porcelain
filtering crucible and transfer the precipitate completely with the aid of O.OlAf-
nitric acid. Wash ten times with 2-cm^ portions of O.OlAf-nitric acid- and five
times with 2-cm^ portions of 95 per cent ethanol. Suck the precipitate as dry as
possible on the pump, dry for 1 hour at 1 10 °C, cool in a desiccator, and weigh as

K2Na[Co(N02)6],H20.

Determination of potassium as potassium tetraphenylboron.*

iscussion. A solution of sodium tetraphenylboron Na'‘'[B(CgH 5 ) 4 ]

This method is only useful where a high degree of agcuracy is not required.

475

XI, 50 QUANTITATIVE INORGANIC ANALYSIS

is probably the best precipitant for potassium, but is expensive. Pre-
cipitation may be cflcctcd at a temperature below 20 ’C in dilute mineral acid
solution (pH 2), in which interference from ino.st foreign ions Is negligible.
The precipitate is granular and settles readily; it is washed with a .saturated
aqueous solution of the precipitate (prepared independently), and the
pota.ssium tetraphenylboron is dried at 120 C and weighed. The compound
decomposes at temperatures above 265 'C, The precipitate is of constant
composition K[fl(C„Hj) 4 ). is sparingly soluble in water (s5,i mg dm"-* of
potassium at 20 ’Q. Very few elements interfere with the determination: these
include the ions of silver, incrcurjfll), thalliumfl). rubidium, and cesiuin;
ammoniuni ion, which forms a slightly soluble salt, can be removed by ignition
prior to the addition of the leagent.

Fivixdiin'. for practice in this determination, weigh out accurately about
O.lOg A.R. potassium chloride and dissolve it in SOenv’ distilled water. Add
lOcin^ of O.l.Q-liydrochloric acid. Then introduce from a burette -lOcm^ of the
sodium tetraphenylboron reagent (1) slowly [5-10 minute.s) and stir
continuously. The temperature throughout must be below 20 C. Allow the
precipitate to settle during I hour. Collect the precipitate on a sintered glass
filtering crucible (poroiity No. 4). wash the precipitate with a .small volume
(5-10 cm^ in small portions] of saturated potassium tetraphenylboron solution
(2), and liiially with l-•2c^H ice-cold distilled water (3). Dry at 120 C and cool
thecoscred crucible in a desiccator. Weigh as KjBfC,, 115 ) 4 ,).

Notes. 1. Prepare tlie sodium tetraphenylboron reagent by dissolving
3.0g of the solid reagent in DOOcm-* distilled water in a glass-stoppercd bottle.
Add about ! g moi.si aluminium hydroxide gel. break up tbe gel if necessary, and
shake the su.spension for 15 minutes, filter through a Wiiatnum No. 40 filter
paper. RefiUer the first part of the filtrate, if necessary, to ensure a clear filtrate.

2. Precipitate about 0. 1 g pulassium (present as poiassium ciiloride in 50cin^
water) with 40 cm^ of the sodium tetraphenylboron solution added slowly and
with constant stirring. Allow to stand for 30 mimiies. filler through a sintered
glass filtering crucible, wasii with distilled water, aiul dry for 1 hour at 120“C.
Shake 20-25 mg of the dry precipitate with 200 cm ^ distilled water in a stoppered
bottle at 5-minutc inters als during I hour. Filter through a Whatman No. 40
filter paper, and u.^e the filtrate as the wash liquid.

3. The reagent is cxpen.sivc; it is therefore desirable to recover it from
potassium TPB pra'ipiiaies remaining from gravimetric determinations or
obtained by adding potassium chloride to filtrates, wash liquids, etc. The
potassium tetraphenylboron is di.s.sulved in acetone and the acetone solution is
passed through a strongly acidic cation c.xchange resin (sodium form): the
diluent contains sodium TPB, and is evaporated to dryness on a water bath. The
resulting sodium tetraphenylboron is recry.stallised from acetone.

C. Determination of polas,sium a.s chloroplatinate ami subsequent weighing a.s
metallic platinum. Ohat^sion. This mciiiod is applicable only to those
potassium compounds whicii can be completely converted into potassium
chloride by evaporation with liydrocliloric acid (as by the technique of Section
XI, 53A), because it is only from a solution containing chloride tiuit potassium
can be completely precipitated a.s K,[PiQ„] by cliloroplatinic acid solution.
Ammonium salts and all metals other than sodium and potassium must be
removed, as must also sulphate, phosphate, and similar radicals. Sodium
chloroplatinate is soluble in 80 per cent ethanol, hence this method provides a

476

GRAVIMETRY XI, 51

means of separation of sodium from potassium. As the composition of the
precipitate may vary slightly from that expressed by the formula K 2 [PtCl 6 ], it is
preferable to convert the potassium chloroplatinate to the metal by reduction
with magnesium ribbon in acid solution, and weigh the platinum. This modified
procedure admits of determining potassium in the presence of sulphates,
phosphates, chlorides, nitrates, borates, sodium, the alkaline-earth metals, iron,
aluminium, and magnesium. Rubidium, cesium, and ammonium salts must be
absent.

Procedure. Weigh out accurately about 0.25 g of the mixed sodium and
potassium chlorides (1) into a small porcelain dish and dissolve it in 5-10 cm^ of
water. Add 5 cm? of hydrochloric acid. Treat with a slight excess of
chloroplatinic acid reagent (2) over that required by the potassium; the presence
of sodium and other salts causes no interference. Crush the precipitate with a
glass rod flattened at one end and collect it in a sintered glass or porcelain filtering
crucible. Wash with ethanol (80-90 per cent by volume) (3), dissolve the
precipitate by pouring hot water over it, and transfer the filtrate and washings
quantitatively to a small beaker. Add 2cm^ concentrated hydrochloric acid,
followed by about 0.5 g magnesium ribbon (previously washed in water) for
every 0.20 g of potassium present. Stir the solution and hold the ribbon at the
bottom of the beaker by means of a glass rod with a flattened end. When the
magnesium has nearly disappeared, add a few cm^ dilute hydrochloric acid, and
allow the platinum to settle. If the reduction is complete, the liquid is clear and
colourless. To make sure add a little more magnesium, and note whether the
solution darkens. Add dilute hydrochloric acid, boil to dissolve any basic salts
which may be present, collect the platinum on a small filter paper, wash with
water until free from chlorides, and ignite in a weighed porcelain or, preferably,
platinum crucible to constant weight. Weigh the platinum, and calculate the
potassium equivalent by the proportion Pt = 2K.

Notes. 1. For practice in this procedure, employ either A.R. potassium
chloride or an artificial mixture of, say, equal weights of A.R. sodium and
potassium chlorides.

2. The chloroplatinic acid reagent is prepared by dissolving 1 g chloroplatinic
acid in 10 cm^ water.

3. All washings must be kept, and the platinum contained in them
subsequently recovered.

XI, 51. SELENIUM AND TELLURIUM. Discussion. The gravimetric
determination depends upon the separation and weighing as elementary
selenium or tellurium (or as tellurium dioxide). Alkali selenites and selenious acid
are reduced in hydrochloric acid solution with sulphur dioxide, hy-
droxylammonium chloride, hydrazinium sulphate or hydrazine hydrate. Alkali
selenates and selenic acid are not reduced by sulphur dioxide alone, but are
readily reduced by a saturated solution of sulphur dioxide in concentrated
hydrochloric acid. In working with selenium it must be remembered that
appreciable amounts of the element may be lost on warming strong hydrochloric
acid solutions of its compounds: if dilute acid solutions (concentration < 6M)
are heated at temperatures below 100 °C the loss is negligible.

With tellurium, precipitation of the element with sulphur dioxide is slow in
ilute hydrochloric acid solution and does not take place at all in the presence of
excess of acid ; moreover, the precipitated element is so finely divided that it

477

XI, 51 QUANTITATIVE INORGANIC ANALYSIS

oxidises readily in the subsequent washing process. Satisfactory results are
obtained by the use of a mixture of sulphur dioxide and hydrazinium chloride as
the reducing agent, and the method is applicable to both tellurites and telluratcs.
Another method utilises e.xcess of sodium hypophosphile in the presence ofdilute
sulphuric acid as the reducing agent.

A process for the gravimetric determination of mixtures of selenium and
tellurium is also described. Selenium and tellurium* occur in practice cither as
the impure elements or as selenides or tcl)uridc.s. They may be brought into
solution by mixing intimately with 2 parts of sodium carbonate and 1 part of
potassium nitrate in a nickel crucible, covering with a layer of the mi.xiure, and
then heating gradually to fusion. The cold melt is extracted with water, and
filtered. The elements are then determined in the filtrate.

A. Determination of selenium. Procedure. The selenium must be present
in the quadrivalent state, and the selenium content of the solution must not
e.xceed 0.25 g per 150cm^. Take an amount of the oxide, selenite, etc., that will
contain not more than 0.25 g selenium, and dissolve it in lOOcm^ concentrated
hydrochloric acid. Add. with constant stirring and at not over 25 'C, 50 cm-* cold
concentrated hydroeliloric acid lliat has been saturated with sulphur dioxide at
room temperature, allow the .solution to stand until the red selenium settles out,
filter Uirough a weighed filtering crucible (Gooch, sintered glass, or porcelain),
wash well successively with cold conccmralcd hydrochloric acid, cold water until
free from chloride, ethanol, and diethyl ether. Dry the precipitate for 3-4 hours
at 30-40 C to remove ether, and then to constant weight at lOO-l 10 C. Weigh
us Se.

B. Dclcrminafion of tellurium. Procedure. The solution should contain
not more than 0.2 g Te in 50citv* of 3.U-hydrochloric acid (c«. 25 per cent by
volume of hydrochloric acid). Heat to boiling, add 15 cm-* of a freshly prepared,
saturated solution of sulphur dioxide, then lOcm* of a 15 per cent aqueous
solution of hydrazinium chloride, and finally 25cm-* more of the saturated
solution of sulphur dioxide. Boil until (he precipitate .settles in an easily filterable,
form ; this should require not more than 5 niinutc.s. .Allow to settle, filter through
a weighed filtering crucible (Gooch, sintered gla.ss, or porcelain), and
immediately wasli witii hot water until free from chloride. Finally wash with
ethanol (to remove all waiter and prevent oxidation), and drv to constant weisht
at 105 X. Weigh as Te.

In the alternative method of reduction, which is particularly valuable for the
determination of small amounts of tellurium, the procedure is a.s follows. Treat
the solution containing, say, up to about 0.01 g To in 90cm-* with lOcm-* of 1 : 3-
sulphuric acid, then add 10 g sodium hypophosphile. and heat on a steam bath
for 3 hours. Collect and weigli tlie precipitated tellurium as abose.

C. Determination of mixtures of selenium and tellurium.

Procedure. Dissolve the mixed oxides (not e.xcccding 0.25 g of each) in
100 cm* of concentrated hydrochloric acid, and add with constant stirring
50cm* cool concentrated hydrochloric acid which has been saturated w'ith
sulphur dioxide at the ordinary temperature. Allow the .solution to stand until the
red selenium has settled, filter through a weighed filtering crucible (Gooch,

* Tellurium ami ils coiiipoumls are toxic and cause itri(,itiun lo cics and skin; contact and inhalation
.should be avoided.

478

GRAVIMETRY XI, 52/53

sintered glass, or porcelain), and complete the determination as described in A.
Preserve the filtrate, hydrochloric acid, and water washings. Concentrate the
latter on a water bath below 100 °C (above 100 °C tellurium is lost as chloride) to
50 cm^ and determine the tellurium as described under B.

XI, 52. SILVER. Determination of silver as chloride. Discussion. The
theory of the process has been given under Chloride (Section XI, 17). Lead,
copper(I), palladium(II), mercury(I), and thallium(I) ions interfere, as do
cyanides and thiosulphates. If a mercury(I) (or copper(I) or thallium(I)) salt is
present, it must be oxidised with concentrated nitric acid before the precipitation
of silver; this process also destroys cyanides and thiosulphates. If lead is present,
the solution must be diluted so that it contains not more than 0.25 g of the
substance in 200 cm^ and the hydrochloric acid must be added very slowly.
Compounds of bismuth and antimony that hydrolyse in the dilute acid medium
used for the complete precipitation of silver must be absent. For possible errors in
the weight of silver chloride due to the action of light, see Section XI, 17.

Procedure. The solution (200 cm^) should contain about 0. 1 g of silver ( 1)
and about 1 per cent by volume of nitric acid. Heat to about 70 °C, and add
approximately 0.2M pure hydrochloric acid slowly and with constantstirring until
no further precipitation occurs ; avoid a large excess of the acid. Do not expose
the precipitate to too much bright light. Warm until the precipitate settles, allow
to cool to about 25 °C, and test the supernatant liquid with a few drops of the acid
to be sure that precipitation is complete. Allow the precipitate to settle in a dark
place for several hours or, preferably, overnight. Pour the supernatant liquid
through a weighed Gooch, sintered glass or porcelain filtering crucible, wash the
precipitate by decantation with O.lM-nitric acid, transfer the precipitate to the
crucible, and wash again with O.OlM-nitric acid until free from chloride. Dry
the precipitate first at 100 °C and then at 130-150 °C, allow to cool in a desic-
cator and weigh. Repeat the heating, etc., until constant weight is obtained (2).
Weigh as AgCl.

Notes. 1. Forexample, from 0.2 g of A.R. silver nitrate.

2. See last footnote in Section XI, 17.

XI, 53. SODIUM. Sodium may be determined in one of the following forms :

A. Determination of sodium as sulphate. Discussion. Any sodium
compound of a volatile acid may be converted into sodium sulphate by repeated
evaporation with sulphuric acid. Some sodium hydrogensulphate is formed in
the process, and this is converted (via the pyrosulphate, Na 2 S 207 ) into the
normal salt with some difficulty. The latter change is facilitated by the addition of
a little powdered ammonium carbonate ; this is because ammonium sulphate,
which is completely volatilised on heating, is formed ;

Na 2 S 207 + (NH4)2C03 = Na 2 S 04 + (NH4)2S04 + CO 2

Procedure. This determination is carried out in a silica or platinum
crucible. Evaporate the solution (I) to dryness in a weighed crucible on the water
bath. Transfer to a triangle (or to a hot plate, the temperature of which can be
controlled), add a few cm^ of concentrated sulphuric acid dropwise, and
evaporate gently to dryness in the fume cupboard until fuming ceases. Repeat t his
operation twice. Allow to cool, add a few small pieces (about the size of a pea) of
solid ammonium carbonate to decompose any pyrosulphate present, and heat to

479

XI, 53 QUANTITATIVE INORGANIC ANALYSIS

dull redness (or at 400-700 ' C) (2) for 1 5 iniiunes. Allow to cool in a desiccator
and weigh the covered crucible immediately it has acquired the laboratory
temperature. Repeat the treatment with atnnioniutn carbonate until constant
w'cight is attained. Weigh as NajS 04 .

Notes, 1. A suitable solution for practice may be prepared by weighing
out accurately about 0.3 g of sodium chloride, and dissolving it in a little water.

2. The temperature should not c-sceed 850'C; any temperature between
400 'C and 700'’C is .satisfactory. Anhydrous .sodium sulphate is .slightly
hygro.scopic.

B. Determination of .sodium as sodium zinc uranyl acetate.*

Discussion. Treatment of a concentrated solution of a sodium salt with a
large e.\ce.ss of zinc uranyl acetate reagent results in the precipitation of
sodium zinc uranyl acetate. This sub.stanccis moderately soluble in water (58.Sg
per lOOOg of water at 20 C) so that a special washing technique must be used.
The solubility in a solution containing e.xcess of the reagent is less. About 10
volumes of the reagent is added for each volume of the sample solution, which
should not contain more than 8 mg of sodium per cm^ ; precipitation of the triple
acetate is usually complete in 1 hour. One mg of sodium yields 66.88 mg of the
triple salt ; the latter is relatively bulky, so that the amount of sodium that can be
handled in a single determination is limited.

Lithium interferes, since it forms a sparingly soluble triple acetate. Potassium
has no elTect provided not more than 50 mg cm' ^ are present. Sulphate must be
absent when potassium is present, for potassium sulphate is .sparingly soluble in
the reagent. Moderate amounts of ammonium salts, calcium, barium, and
magnesium may be tolerated; for larger amounts, a double precipitation is
ncce.ssary. Phospliates. arsenates, molybdates, oxalates, tartrates, sulphates (in
the presence of potassium), and strontium interfere.

Procedure. The neutral or feebly acid sample solution, free from the
interfering substances mentioned abosc, should contain not more than 8 mg of
sodium perem^, preferably as chloride. Treat the sample solution (say, l.5cin^)
( 1) with I .Sciit’ of zinc uranyl acetate reagent (2), and stir vigoromsly, preferably
mechanically for at least 30 minutes. Allow to stand for 1 hour, and filter through
a weighed sintered glass or porcelain filteringcrucible (porosity .No. 4). Wash the
precipitate four limes with Z-env* portions of the precipitating reagent (allow the
wash liquid to drain completely before adding the nc.xt portion), then ten times
with 95 per cent ethanol saturated with sodium zinc uranyl acetate at room
temperature (2-cm-’ portions), and (inally with a little dry diethyl ether or
acetone. Dry for 30 minutes oulv at 55-60 C (3). Weigh as
NaZn(UO,)j(C,HjOj)„.6H,0.

Notes. 1 . A suitable solution for practice may be prepared by evaporating
20.0cm^ of 0.02A/-sodium chloride, prepared from the A.R. salt, to 1.5 cm^ on a
water bath.

2. The reagent is prepared by mixing equal volumes of solutions A and B and
filtering after standing overnight.

Solution A: dissolve 20g crystulli.scd uranyl acetate U02(C,Hj,0;)22l-i20 in
4 cm glacial acetic acid and lOOcni'^ water (warming may be necessary).

This method is oitly of use where higli degree of accuracy is not required.

480

GRAVIMETRY XI, 54

Solution B: dissolve 60 g crystallised zinc acetate Zn(C 2 H 302 ) 2 , 3 H 20 in S cm^

glacial acetic acid and 100 cm^ of water. .

3. Alternatively, draw air through the crucible for 5 nainutes to volatilise the
solvent, wipe off any condensed moisture on the outside with a clean linen cloth,
allow to stand in the air or in a desiccator for 10-1 5 minutes, and weigh.

XI, 54. STRONTIUM. Strontium may be determined in one of the following
forms :

A. Determination of strontium as sulphate. Discussion. In this de-
termination (probably the most accurate for strontium) calcium, barium, and
lead must be absent, and the solution (preferably of the chloride) should be
nearly neutral. If considerable quantities of acid are present, this must be
removed by evaporation. Strontium sulphate dissolves appreciably in an acid
medium because of the reaction :

SrS 04 +H+ ^HS 04 - +Sr2 +

Strontium sulphate has a solubility of about 0.14 g dm“^ at the laboratory
temperature; the solubility is decreased by the addition of a slight excess of
sulphuric acid, and of ethanol (50 per cent).

Procedure. The solution (100 cm^) should contain about 0.2 g of strontium
and be very slightly acid with hydrochloric acid (1). Add slowly a ten-fold excess
of dilute sulphuric acid, followed by a volume of ethanol equal to that of the
solution. Stir well, and allow to stand for at least 12 hours. Transfer the
precipitate to a weighed Vitreosil or porcelain filtering crucible, wash with 75 per
cent ethanol to which a few drops of sulphuric acid have been added, and finally
with pure ethanol until the washings are free from sulphate. Dry and ignite
(crucible ignition-dish or in a large nickel crucible) at dull redness (or in an
electric muffle furnace at 500-600 °C) to constant weight. Alternatively, a filter
paper may be used ; here the paper should be burnt apart from the precipitate (to
prevent possible reduction of the latter to the sulphide), and the residue then
ignited together with the main precipitate in a weighed porcelain, silica, or
platinum crucible. Weigh as SrS 04 .

Note. 1 . A solution for practice in this determination may be prepared by
dissolving 0.3-0.4g, accurately weighed, of pure strontium carbonate in a little
dilute hydrochloric acid (see Section XI, 20), and diluting to 100 cm^.

B. Determination of strontium as strontium hydrogen phosphate,
SrHP 04 . Discussion. Strontium (30-200 mg) may be precipitated as SrHP 04
using potassium dihydrogen phosphate; precipitation commences at pH 4 and is
quantitative at pH 5.7-6. The flocculent precipitate soon becomes crystalline. It
may be weighed as SrHP 04 after drying at 1 20 °C ; alternatively, it may be ignited
and weighed as Sr 2 P 207 . Ions which yield insoluble phosphates should be
absent; the sodium, potassium, or ammonium ion concentration should not
exceed 0 . 2 Af.

Procedure. Treat the sample solution (60 cm^ ; say, prepared by weighing
out accurately about 0.15 g pure SrCl 2 , 6 H 20 and dissolving in water) with
AHa ^ O-^Jlf-potassium dihydrogen phosphate and heat to the boiling point.
Add IM-potassium hydroxide from a dropper pipette until an appreciable
is formed. The final pH should be about 6 ; this can be detected by
adding the base until bromocresol purple indicator in the solution just turns
purple, or a pH meter may be used. Boil until the initial flocculent precipitate

481

XI, 55 QUANTITATIVE INORGANIC ANALYSIS

becomes crystalline (30-60 minutes). Allow to sUmd for 1 hour. Collect the
precipitate in a sintered glass (porosity No. 4) or porcelain filtering crucible;
remove any precipitate adhering to the walls of the beaker with a rubber-tipped
stirring rod, and wash the precipitate with a little cold water. Dry at 120“'C.
Weigh as SrllPOj.

XI, 55. THALLIUM. Thallium may be determined in either of the following
forms;

A. Delcrmination of thallium as chromate. Discussion. The thallium*
must be present in the thallium(l) state. If present as a thallium(lll) salt,
reduction must be effected (bcforcprecipitation) with sulphur dio.xide: the c.xcess
of sulphur dio.xidc is boiled off.

Proietiurc. The solution (lOOcm^) should contain about 0.1 g of Tl, no
excessive amounts of ammonium salts, and no substances that form precipitates
with ammonia solution, or reduce potassium chromate, or react with potassium
or thalliumlD chromate in animoniacal solution. Neutralise the thallium
solution with dilute ammonia solution (211), and add 3cm^ in excess. Heat to
about SO C, and add 2g of potas.dum chromate in the form of u 11) percent
solution slowly and with constant .stirring. Allow to stand at the laboratorj'
temperature for at least 12 iiours. Filler through a weighed filtering crucible
(Gooch, sintered glass, or porcelain), wash with I percent potassium chromate
solution, then sparingly with 50 per cent ethanol, and dry at 120'"C to constant
weight. Weigh as Tl .CrO^.

B, Deterutination of thallium with tetraphcnylarsoniuin chloride.

Discussion. Thaliium((n) in excess of hydrochloric acid reacts with tetra-
phenylarsonium chloride [(QHjl^AslCl to give an insoluble tetraphenylar-
.sonium chlorothallate :

{QHj) 4 A.s]'+[Tia,]- - ((QH,),As)(TlCI,l

I'he precipitate can be dried at 1 lO'C. Thallium(l) is readily o.xidi.sed to
tlialliumdil) by hydrogen peroxide in alkaline solution. The principal
interferences are cations which form insoluble chlorides and also various anions
(c.g., lluoride, iodide, bromide, thiocyanate, nitrate, perchlorate, periodate,
permanganate, per-rhenaie. molybdate, chromate, and tungstate). The
precipitate must be washed with dilute hydrochloric acid, otherwise hydrolysis
occurs and low rc.sults arc obtained.

Proceilnrc. Oxidise thailium(I) in llic sample solution {75cm^ containing
up to 90 mg of thallium) (1) to ihallium(ill) by adding 2.()cm'’ TOO-volunie'
hydrogen peroxide in the presence of sodium hydroxide solution. .Acidify with
hydrochloric acid and add a few cm^ concentrated hydrochloric acid in e.xcess; a
white precipitate, probably a Tl{l)~Tl(nij complex, fornus, but this will dissolve
upon the addition of a further 1 cm^ of 30 per cent hydrogen peroxide to the acid
solution. Dilute the .solution to render it 0.5-2.0.I/ in hydrochloric acid, and add
excess of the reagent solution (2). Heat to boiling to coagulate tlie white
precipitate and keep overnight. Collect the precipitate on a weighed sintered
glass filtering crucible (porosity No, 4). wash with 20-40 cm^ L)/-hydrochloric
acid, and dry at 1 10 •'C. Weigh as ((Ce.Hj).,As)[TlCl.t].

• JlKillium and it;, compounds are io,\ic and cause irritation to cses and the skin; conlacl and
inhalation should be avoided.

482

GRAVIMETRY XI, 56

Notes. 1 . A solution of 0.05-0. 1 0 g, accurately weighed, A.R. thallium(I)
sulphate in 50-75 cm^ water may be used for practice in this determination.

2. The reagent solution is, prepared by dissolving 6.7 g [(C 6 H 5 ) 4 As]Cl in
lOOcm^ water. Ten cm^ suffice for the .precipitation, of 90mg of thallium.
Tetraphenylarsonium : chloride ' is' available from the G. Frederick Smith
Chemical Go., Columbus, Ohio, and from Fluka A.G., Buchs, Switzerland.

XI, 56. THORIUM. Thorium may be determined in either of the following
forms:

A. Determination of thorium as sebacate and subsequent ignition to the oxide,

ThOj. Discussion. This procedure permits of the separation by a single
precipitation of thorium from relatively large amounts of the lanthanoids (Ce,
La, Pr, Nd, Sm, Gd) and also from quadrivalent cerium. •

Procedure. The solution (100 cm^) should be neutral or faintly acid, and
contain not more than 0. 1 g Th. Heat the solution to boiling and add slowly and
with constant stirring a hot almost saturated solution of pure sebacic acid in
slight excess. The precipitate is voluminous, but granular, and therefore easily
manipulated. Filter off immediately, wash thoroughly with hot water, dry, and
ignite (use either a Meker or Fisher burner or an electric muffle furnace at
700-800 °C) in a weighed platinum, porcelain, or silica crucible to constant
weight. Weigh as ThOj.

B. Determination of thorium as iodate, and subsequent ignition to tbe oxide,
ThOj, via the oxalate. Discussion. Thorium iodate is precipitated quanti-
tatively by potassium iodate from nitric acid solution: a separation from the
lanthanoids, trivalent cerium, iron, aluminium, and phosphoric acid is thus
achieved. Titanium, zirconium, and cerium(IV) accompany thorium, and must
therefore be absent. The thorium iodate is dissolved in hydrochloric acid,
precipitated as the oxalate, and ignited to and weighed as Th 02 .

Procedure. The solution (100 cm^) should be chloride-free (1) emd contain
not more than 0.2 g Th. Add 50 cm^ concentrated nitric acid and cool in ice
water. Add a cold solution of 15 g A.R. potassium iodate in 30 cm^ water and
50 cm^ concentrated nitric acid; stir occasionally during 30 minutes. Allow to
settle, break up any lumps of precipitate with a glass rod flattened at one end,
filter through a hardened quantitative filter paper, wash with 250 cm^ of a cold
solution containing 8 g potassium iodate and 200 cm^ dilute nitric acid ( K 1 ) per
dm^, and allow to drain. Transfer the precipitate back into the original beaker
with the aid of lOOcm^ of the wash solution, stir thoroughly and filter through
the same filter paper. Allow to drain and again transfer the precipitate back into
the beaker, but this time with a little hot water. Heat nearly to boiling, and
dissolve the precipitate by adding 30 cm^ of concentrated nitric acid slowly and
with stirring. Dilute to 60-100 cm^, and re-precipitate the thorium as iodate by
adding a solution of 4g A.R. potassium iodate dissolved in a little hot water
acidified with nitric acid: allow to cool. Filter, wash by decantation as before
with 100 cm^ of the wash solution, and transfer the precipitate to the paper.

To remove any titanium, zirconium, or ceriumQV) which may be present,
place the filter and precipitate in the original beaker and dissolve the precipitate
by boiling with hot dilute hydrochloric acid and a little sulphurous acid. Dilute,
precipitate with ammonia solution, filter, and wash'the precipitate with hot water
until free from iodides. Dissolve the precipitate again in hydrochloric acid, and
precipitate the thorium as oxalate by adding slowly and with constant stirring

483

XI, 57 QUANTITATIVE INORGANIC ANALYSIS

sutlicienl of a boiling 10 per cent solution of oxalic acid to combine with all the
thorium and leave an excess of 20cm^. Allow the solution to cool and stand
overnight. Filter through a quantitative filter paper, wash with a solution
containing 3.5cin-* concentrated liydrochloric acid and 2.5 g oxalic acid per
100cm-'. Ignite the precipitate as in method A above. Weigh as ThO,.

Note. 1. The solution may contain sulphuric acid; for example, that
obtained by dissolving monazitc sand in sulphuric acid,

XI, 57. TIN. Tin m;iy be determined in any of the following ways,

A. Determination of tin with cupferron and weighing as tin{IV) oxide,
SnOj. Disatsxkm. This process permits of the precipitation of tin in the
presence of aluminium, chromium, cobalt, nickel, and manganese. In the
presence of 5 car' oNS percent hydrotluoricacid per300em^ of solution which is
ai. 0.2.1/ in hydrochloric acid, tin(lV) (about 0. 15g) is not precipitated by
hydrogen sulphide, whereas copper, lead, arsenicdll) and antimony(in) are
precipitated; :tfter adding about .3 g boric acid, and boiling to expel hydrogen
sulpiiide, the tin is precipitated in the filtrate by cupferron.

/Yoret/nre. Remove metals such as copper, lead, Irivaleni arsenic, and
antimony, if present, by precipitation with hydrogen sulphide in the presence of
hydrolluoric acid (see Di.\cus.sia/i). The solution should contain about 0.15 g tin
and occupy a volume of 250-300 cm^. Add 3g boric acid, boil off the e.xcess of
hydrogen sulphide, introduce 2.5 cm^ concentrated sulphuric acid cautiously,
followed by a liberal excess of a filtered 10 jx'r cent aqueous solution of
cupferron. Stir vigorously (cm 30---I(J minutes) until the precipitate becomes
compact and brittle; it may then be crushed to a fine powder with a glass rod.
Filter upon a Whatman No. -I I or 5-41 filter paper, wash with cold water, dry in a
weighed crucible, expel the organic matter by gentle ignition, and then ignite to
constant weight. Weigh as SnO,.

B. Determination of tin with iV-bciuoj I-/V-pheii> Ihydro.\vIaminc.
D/.scu.vs/wj. A'-Bcn/.oy!-jV-phcnylliydroxylamine C„HjCON(OH)CoHj, as a
1 per cent solution in ethanol, preeipitafc.s a complex (C,jH,,OiN)jSnCL,
lu.p. 171 'C, from tin(l\') .solutions containing I -8 per cent concentnitcd
hydrochloric acid; the complex can be dried at 1 10'"C. Apparently the reagent
reduces tin(IV) to tin(ll) and then forms the addition compound. Copper can be
qu;intiiatively precipitated by the re;igent at pH 3.6-6.(): no interference is
encountered Irom copper, lead, or zinc in precipitating tin from, for e.xamplc,
brass solutions containing 7 per cent by volume of concentrated hydrochloric
acid.

A'-Benzoyl-A'-pheuylhydroxylaminc is a wiiite crystalline solid, m.p. 121 °C;
its solubility in water is 0.04 g per lOOcm^ at 25 C and 0.5 e per lOOcm^ at about
80 ’C.

The reagent has been used for the determination of copper, iron, and
aluminium. The pH range.s for quantitative precipitation are; Cu, 3. 6-6.0; Fe,
3.0-5. 5; and Al, 3. 6-6.4, Incomplete precipitation occur.sat a lowcrpH.and high
results arc obtained at higher pH values. Titanium must be precipitated below
25 'C and ignited to, and weighed as. the dioxide. Zirconium is also precipitated.
Iron and aluminium cannot be precipitated in the presence of phosphate;
ehromiuni(lll) interlercs with the precipitation of ironflll). The following
elements do not give precipitates with the reagent at pH 4; bismuth, cadmium,
cobalt, manganese, nickel, uranium(lV). and zinc.

484

GRAVIMETRY XI, 58

Procedure. To the sample solution of tin(IV) chloride (containing 5-20 mg
of tin), add 10 cm^ concentrated hydrochloric acid and dilute to ca. 600 cm^ with
distilled water. Add from a separatory funnel, dropwise and with constant
stirring, 5 cm^ of a 1 per cent solution of the reagent in ethanol for each 10 mg of
tin present plus 8 cm^ in excess. Cool in an ice bath for 4 hours, filter on a filtering
crucible (sintered glass or porcelain), wash with a few cm^ of ice water, and dry at
110°C. Weigh as (Ci3Hii02N)2SnCl2.

XI, 58. TITANIUM. Titanium may be determined in one of the following
forms. .

A. Determination of titanium, with tannic acid and phenazone.

Discussion. This method affords a separation from iron, aluminium,
chromium, manganese, nickel, cobalt, and zinc, and is applicable in the presence
of phosphates and silicates. Small quantities of titanium (2-50 mg) may be
readily determined.

Procedure. The Ti content of the solution should not exceed 0. 1 g of Ti02,
and the titanium should be present as the sulphate or chloride. Add dilute
ammonia to the solution until the odour persists, then (cautiously) lOcm^
concentrated sulphuric acid and 40 cm^ of 10 per cent tannic acid solution. Dilute
to 400 cm^, stir thoroughly, and cool. Introduce a 20 per cent aqueous solution of
‘phenazone’ (antipyrine) 2,3-dimethyl-l-phenyl-5-pyrazatone with constant
stirring until an orange-red flocculent precipitate is obtained. Stop the stirring,
and continue the addition of the phenazone solution until a white, cheese-like
precipitate (produced by the interaction of tannic acid and phenazone) is formed
in addition to the red precipitate. Boil the mixture, remove the flame, add 40 g
ammonium sulphate, and allow to cool with occasional stirring. Filter the bulky
precipitate through a Whatman No. 41 or 541 filter paper, supported on a
Whatman filter cone (hardened. No. 51), with slight suction, and wash with a
solution of 100 cm^ water, 3 cm^ concentrated sulphuric acid, 10 g ammonium
sulphate, and 1 g phenazone. Dry the precipitate at 100 °C, transfer to a weighed
crucible, heat gently at first, and then ignite at 700-800 °C to constant weight.
Weigh as TiOj.

Note. If the wet precipitate is heated directly, caking occurs which renders
the complete oxidation of the carbonaceous matter very slow. If alkali metals
were originally present, the ignited oxide must be washed with hot water, filtered,
and re-ignited to constant weight.

B. Determination of titanium with 4-hydroxyphenylarsonic acid.

Discussion. This procedure will separate titanium from most other com-
monly occurring ions by a single precipitation. Zirconium, tin, cerium(lV),
and hydrogen peroxide must be absent.

Procedure. Dissolve the sample ( 1 ) containing not more than about 0.06 g
TiOj in sulphuric or hydrochloric acid, and dilute to 200 cm^. The amount of the
acid present should be such that the solution will be approximately, but not more
than, 0.6M in hydrochloric acid or 0.9M in sulphuric acid after the reagents have
been added and the precipitation is complete. Heat the solution to boiling ; if iron
is present, add 2-3g A.R. ammonium thiocyanate: add 100 cm^ of a 4 per cent
aqueous solution of 4-hydroxyphenylarsonic acid, H 0 -C 6 H 4 -As 03 H 2 .. Boil
gently for 15 minutes to coagulate the precipitate. Allow to cool to room
temperature, and filter with suction on a Whatman No. 542 or 42 filter paper
supported on on a filter cone (Whatman, No. 5 1 , hardened). Wash the precipitate

485

XI, 59 QUANTITATIVE INORGANIC ANALYSIS

live or six times with a wash liquid of 0.25A'-hydrochloric or sulphuric acid
containing about 0.5g of the solid reagent per 100cm-* (if iron is present, l-2g
ammonium thiocyanate should be also added to each 100 cm^ of wash liquor).
Finally, wash the precipitate two or three times with 2 per cent aqueous
ammonium nitrate solution. Transfer the filler to a .silica crucible, ignite gently at
first until all the carbon is burnt olf (this operation must be curried out in a fume
clumiher (hood) provided with an efficient draushi) and then with a Fisher burner
or in an electric inutile furnace at 700-800 C until constant weight is attained.
Weigh asTiOj.

Note. I. For practice in this determination the Bureau of Analysed
Samples ‘Iron Ore, No. 175' may be used. Dissolve 4g of this in 100cm* dilute
hydrocliloric acid and tiller. Fuse the nndissolvcd residue with sodium
carbonate, wash the melt into the main iiitrate, remove the silica in the usual
manner, add 4g A.R. ammonium thiocyanate, dilute to 20{)-250cm\ and
continue the determination ;is above.

XI, 59. TUNGSTKN. Tungsten. a.s tungstate, may be determined in one of
the following forms.

A. Determination of tungsten as the trio.vide (tannic acid- phenazone
method). Discussion. Tungstic acid is incompletely precipitated from
solutioms of tungstates by tannic acid. If, howeser, phenazone (2,3-dimethyI-l-
phenyl-5-pyra-/.aione) is added to the cold solution after treatment with c.xcess of
tannic acid, precipitation is quantitative. This proce.ss e/rccis a separation
from aluminium, and tdsu from iron, chromium, mangane.se. zinc, cobalt, and
nickel if a double precipitation is used.

Procedure. The solution of tungstate {200~250cm*) should contain not
more than 0.15g of WOj, and be faintly ammoniacal. Add 6-7cm* of
concentrated sulphuric acid and 7 -Sg of ammonium sulphate, and heat to
boiling. Treat with 6 enr’ of 10 per cent aqueous tannic acid solution, keep the
mi.xture on the water batii for a few minutes, and allow to cool to room
temperature. A fltKculent dark-brown precipitate separates. When cold, stir in
10cm* of a 10 per cent a<iucous solution of phenazone. Filler the precipitate
through a weighed silica, Gooch, or porcelain filtering crucible ( 1 ), wash with the
special wash liquid (2), and ignite to constant weight at 800-900‘'C. Weigh as
WO,.

.Notes. 1. The filtrate must be colouric.ss. If it is yellow, insuHicient
phenazone has been added.

2. The special wash liquid contains I cm* concentrated sulphuric acid, lOg
ammonium sulphate, and 0.4 g phenazone in 200cm* of water.

B. Detcrinination of tungsten as hurium tungstate. Discussion. A dilute
neutral solution of a tungstate (pH about 7.7) is precipitated by barium chloride
solution as barium tungstate. It is important that the solution be dilute; in
concentrated solutions high results are obtained, due to co-precipitation of
barium chloride. In extremely dilute solutions and at low temperatures, a fine
precipitate is slowly formed, which tends to pass through the filter and adhere to
the walls of the beaker. The solubility of barium tungstate is 4 mg dm “ * at 22 C,
and 0.02 mg dm”* in the pre.sence of a 50 per cent excess of barium chloride; it
increases rapidly with decreasing pH.

Procedure. The solution of tungstate (250 cm*) may contain about 0.15g
of W (1) and be almost neutral (pH 7-8). Adjust the pH of the solution, if

486

GRAVIMETRY XI, 60

necessary, by the addition of dilute acetic acid or of dilute sodium hydroxide
solution. Heat to boiling and add a saturated solution of crystallised barium
chloride in lOcm^ of water dropwise and with constant stirring. Allow the
suspension to stand to acquire the laboratory temperature, filter through a
porcelain filtering crucible, wash with cold water until the washings are free from
chloride, dry at about 750 °C (in an electric crucible furnace) to constant weight.
Weigh as BaWO^. . ■

Note. 1. Use a solution prepared from about 0.25 g A.R. sodium tungstate,
Na2W04,2H20 (accurately weighed), for practice in this determination.

XI, 60. URANIUM. , U ranium, as uranyl salts, may be determined in either of
the following forms.

A. Determination of uranium >vith oxine. Discussion. The formula of the
compound is noteworthy, for it differs from all other metallic ‘oxinates’ (compare
Section XI, IIC). This method may also be employed for the titrimetric
determination of uranium with standard potassium bromate solution (compare
Section X, 140).

Procedure. The uranium should be present as uranyl nitrate or chloride in
1-2 per cent acetic acid solution (1); up to 0.3 g of .U may be present in 200 cm^ of
solution. Add 5 g A.R. ammonium acetate, heat to boiling, and add 4 per cent
oxine solution (2) dropwise and with stirring : use 0.5 cm^ of the reagent for every
lOmg of U present and a further 4-5 cm^. Heat on a boiling water bath for 5-10
minutes. Allow to cool, and filter through a sintered glass or porcelain fiiltering
crucible; wash several times with hot water and then with cold water. Dry to
constant weight at 105-1 10 °C, and weigh as U02(C9H60N)2’C9H70N.
Alternatively, the precipitate may be ignited to and weighed as UaOg.

Notes. 1. If the solution contains mineral acid, almost neutralise with
ammonia solution (or add dilute ammonia solution until a faint turbidity persists
and render the solution just clear with a few drops of dilute hydrochloric acid),
add 5 g A.R. ammonium acetate and then sufficient acetic acid to give a 1-2 per
cent solution.

2. Details of the oxine solution are given under Section XI, 46B.

B. Determination of uranium with cupferron. Discussion. Cupferron
does not react with hexavalent uranium, but tetravalent uranium is
quantitatively precipitated. These facts are utilised in the separation of iron,
vanadium, titanium, and zirconium from uranium(VI). After precipitation of
these elements in acid solution with cupferron, the uranium in the filtrate is
reduced to the tetravalent state by means of a Jones redactor and then
precipitated with cupferron (thus separating it from aluminium, chromium,
manganese, zinc, and phosphate). Ignition of the uranium(IV) cupferron'
complex affords UjOg.

Procedure. If uranium is to be determined in the filtrate from the
precipitation of the iron group by cupferron, concentrate the solution to 50 cm^,>
add 20 cm^ of concentrated nitric acid and 1 0 cm^ of concentrated sulphuric acicl
(if not already present) and evaporate until fumes of sulphur tri oxide appear. If
organic matter still remains (as shown by the appearance of a dark colour upon
evaporation), repeat the treatment with nitric acid. Finally, expel the nitric acid
by evaporating to strong fuming, after the addition of a little water. Dilute the
mo that it contains about 6cm^ of concentrated sulphuric acid per

100 cm . Cool to room temperature and pass the solution through a Jones

487

XI, 61/62 QUANTITATIVE INORGANIC ANALYSIS

rcducior (Section X, W3j; wash the rcductor with 5 percent sulphuric add, cool
the combined reduced solution and washings to 5-10 'C. and add excess of a
freshly prepared 6 per cent solution of cupferron. The precipitate does not
usually form until about 5cm^ cupferron solution has been added. Introduce a
Whatman 'accelerator' or one-quarter of an 'ashless tablet’, allow to settle fora
few minutes, and filter through a quantitative filter paper. Wash with cold 4 per
cent sulphuric acid containing 1 . 5 g of cupferron per dm’- Dry the precipitate at
100 'C, ignite cautiously in a platinum crucible, lirsi at a low temperature and
then at 1000 C. to constant weight. Weigh as UjO„.

XT, 61. V.VN.ADIUM, This element, as vanadate, may be determined in the
following form.

Determination of taiutdium as siher vanadate. DLsemsian. Vanadates are
precipitated by e.xccssofsilver nitrate solution in tlie presence of sodium acetate;
after boiling, the precipitate consists of silver orthovanadate. The following
reactions occur with a .solution of a inctavanadatc:

2NaVO, -iOCHjCOONa -i- H.O Na^V.O- + 2C1 liCOOH
NhuV.O- rOAuNO, ce:.-\g,;V,0, 4-4N'aN03
V,0, f 2 AgN 63 -f 2CH jCOONa + i 1 ,b

2 Ag j VO^ f- 2 CH jCOOH -r 2NaNOj

Tiirimeliic methods (see Chapter X) are, however, more convenient, less
influenced by interfering elemems. and are generally preferred.

ProMlun’. Neutralise ilic solution (200cm-*), containing not more than
0.2 g of alkali vanadate, if acid, by aqueous sodium hydro.xide. or, if alkaline, by
the addition of nitric acid to the boiling solution until it becomes yellow, followed
by decoloiisation with dilute ammonia solution. Add 3 g of ammonium acetate,
0.5cm’ of concentrated ammonia solution, and then e.xcess of silver nitrate
solution, heat to boiling and then keep on a steam bath for 30 minutes. Test for
complete precipitation with more silver nitrate solution; if a turbidity is
produced, boil the liquid until it becomes clear. Allow the dense brown
precipitate of silver vanadate to settle, and collect it on a weighed filtering
crucible (Cunxih. smtereti class, or porcelain), wasii with hot water, and dry at
110 C. Weigh as AgjVO^.

It has been staled lliul the results obtained by precipitation of vanadate as
silver oriho-vanadate Ag, VO 4 are not altogether satisfactory. Better results are
obtained by precipitation at pi! 4.5 as silver mela-vanadale AgVOj; the
precipitate is weighed after drying at 100 - 105 C.

XI, 62. ZINC Zinc may be delennincil in any of the following forms.

A. Determination of zinc as quinuldate. Disat.w'cm. Quinaldicacid orits
sodium salt precipitates zinc quantitatively from dilute acetic acid or slightly
ammoniacal solutions. Iron, aluminium, chromium, beryllium, titanium, and
uranium interfere in acid solution, hut in the presence ofalkali tartrate in alkaline
solution only zinc precipitates; copper and cadmium must be absent. The reagent
is described in Section XI, 1 IM.

Procedure. The solution may contain not more tlian 0. 1 g Zn, and should
be acidified with 2-5 cm* acetic acid (to pll 3-4). ileal to boiling and add 3 per
cent sodium quinaldate solution with stirring until precipitation is complete; an
excess ol 25 per cent should be usetl. Allow to cool to room temperature. Wash

488

GRAVIMETRY XI, 62

the precipitate by decantation with cold water, collect it on a sintered glass or
porcelain filtering crucible, wash with a little ethanol, and dry at 105-1 10 °C to

constant weight. Weigh as Zn(CioH 602 N) 2 ,H 20 . . .

B. Determination of zinc by the pyridine method. Discussion." This
method is a very rapid one, but unless the various wash solutions are carefully
prepared, low results will be obtained. The complex may be kept unchanged in a
vacuum desiccator for 2-3 hours (see Section XI, IIK). Large quantities of
ammonium salts must not be present, as these exert a slight solvent action upon
the precipitate. If the solution is strongly acid, it must be evaporated to dryness
and the residue dissolved in water.

Procedure. The solution (75 cm^) should contain about 0.05 g zinc ( 1) and
be neutral or very faintly acid. To the cold solution add 1 g solid A.R. ammonium
thiocyanate, followed by 1 cm^ pure pyridine. Shake vigorously, when a white
crystalline precipitate will separate. (Precipitation may also be carried out in hot
solution; the complex separates in comparatively large crystals on cooling.)
Allow to stand for 15 minutes, and stir frequently. Filter through a weighed
sintered glass or porcelain filtering crucible, and transfer the precipitate to the
crucible with the aid of Solution 1. Wash the precipitate four times with Solution
2, then wash the walls of the crucible with 1 -cm^ portions of Solution 3 (use a 1 - or
2 -cm^ pipette for this process), and finally five to six times with 1-2 cm^- volumes
of Solution 4. It is important to suck well on the pump between each washing; it is
also advantageous to stir the precipitate with a thin glass rod when washing with
Solutions 3 and 4. Dry the crucible and precipitate in a vacuum desiccator for 1 5
minutes, and weigh. Repeat the drying process until the weight is constant.
Weigh as [Zn(C 5 H 5 N) 2 ](SCN) 2 .

Solution 1. 100 cm^ water containing 0.3 g NH 4 SCN and 0.5 cm^ pyridine.

Solution 2.85.5 cm^ water, 1 3 cm^ of 95 per cent ethanol, 0. 1 g NH 4 SCN, and
1.5 cm^ pyridine.

Solution 3. 1 0 cm^ absolute ethanol + 1 cm^ pyridine.

Solution 4. 15 cm^ diethyl ether (sodium dried) + 2 drops pyridine.

If the wash solutions have been prepared, the determination should be
completed within an hour.

Note. 1. For practice in this determination, employ about - 0.25 g,
accurately weighed, A.R. zinc sulphate, or about 0.6 g pure ammonium zinc
sulphate (NH 4 ) 2 S 04 -ZnS 04 , 6 H 20 ; prepared by mixing equimolecular amounts
of A.R. zinc sulphate and A.R. ammonium sulphate dissolved in boiling water,
and re-crystallising the product twice from hot water. The crystals are air dried,
and 0.6 g, accurately weighed, is dissolved in 75 cm^ water.

C. Determination of zinc as 8 -hydroxyquinaldinate. Discussion. Zinc
may be precipitated by 8 -hydroxyquinaldine ( 2 -methyloxine) in acetic
acid-acetate solution : it can thus be separated from aluminium and magnesium
(see Section XI, IIP). It can be weighed as Zn(CioH 80 N )2 after drying at
130-140 °C. The co-precipitated reagent is volatile at 130 °C.

Procedure. The solution may contain up to 0.05 g Zn in 200 cm^. Add
dilute aqueous ammonia solution until a white precipitate of zinc hydroxide just
appears. Re-dissolve the zinc hydroxide with a drop of acetic acid. Add a slight
excess of the reagent (1) (2 cm^ for each 10 mg of Zn present) and then 2-3 drops
ot concentrated ammonia solution; the pH should be at least 5.5. Digest the
precipitate at 60-80 °C for 15 minutes, allow to stand for 10-20 minutes, and
filter through a sintered glass or porcelain filtering crucible. Dry to constant

489

XI, 63 QUANTITATIVE INORGANIC ANALYSIS

weight at 130-I-I0“C. Weigh as Zn(CioH^ON)j.

If aluminium is present, acid 1 g of ammonium tartrate to the clear, slightly
acid solution. Introduce the reagent (2 cm^ for each 10 mg of Zn present), dilute
the solution to 200cm-*, and heal to 60-80 C. Neutralise the e.xcess of acid by
adding dilute ammonia solution (llS) dropsvise until the complc.x salt which
forms on the addition of each drop just rc-dissolvcs on stirring. Add, with
stirring, 45cin^ of 2,\/-aimnonium acetate solution. The pH should be at least
5.5. Allow the solution to stand for 10-20 minutes, and complete the
deiennination as above.

Note. 1 . The reagent i.s prepared by dissolving 5 g 8-hydroxyquinaldinc in
12g glacial acetic acid and diluting to lOOcm^ with water.

XI, 63. ZlRCO.NiUM. Zirconium may be determined in one of the following
forni.s.

A. Delerniinatioii of zirconium vriih sclenlous acid, and .subsequent ignition to
the dio.xide, ZrO^. Disaix'iinn. The zirconium i.s precipitated as the b;uic
selenite with selenious acid in hot dilute hydrocliloric acid solution, the
precipitate washed witii dilute hydrochloric acid, and then ignited to, and
weighed as, ZrOj. No other acids should be present, and the hydriKbloric acid
content should preferably be 5 per cent and not over 7 per cent by volume. This
nietluHl enables a separation to be clfccted by a single precipitation from the rare
earths (cerium being in the irivaknt condition) and from aluminium. Iron, if
present up to 10 per cent of the weight of the zirconium, docs not interfere if
precipitation is made m dilute solution (lOO-lOOcm"' of solution containing
0.05 g ZrOj) and double the quantity of selenious acid solution is used for
precipitation.

In an aliernaiive meihod (he zirconium i.s converted into the normal .selenite,
ZrlSoOj),, after digestion at 8() -10l) 'C and weighed as such after drying at
110-150 C.

FrocaJuri’. The solution (200cm^) .should coniaiii about 5 per cent by
volume of hYdroehloric acid (sulphuric acid is undesirable) and not more than
0.2g zirconium (as ZrO,). Treat with 20cin-* of 12.5 per cent aqueous selenious
acid solution, and boil for a few minutes. .Allow the precipitate ofbasic selenite to
settle, filter ihrougiv a quantitative filter paper, wash with hot 3 per cent
hydrochloric acid containing a little seleniou.s acid, dry in a weighed porcelain, or
platinum crucible, and ignite (a temperature of 900-1000 C is satisfactory) to
constant weight. Weigh as ZrO,.

Alternatively, precipitate the zirconium as the normal selenite, filter through a
sintered glass filtering crucible (porosity No. 4). wa.sli with hot dilute
hydrochloric acid, followed by cold water until the washings are free from
selenious acid, and dry at 110-150 C to constant weight. The precipitate
contains 26.43 per cent Zr. Some reduction may occur with large quantities of
precipitate, leading to slightly high results.

B. Deleriiiinalion of zirconium with niandelic acid and .subsequent ignition to
the dioxide, ZrOj. 1) Ac ms ion. Zirconium may be precipitated from a
hydrochloric acid solution with mandclic acid (C„H 5 -CH{OH)-COOH) as
zirconium mandelatc, Zr(CsIl 70 j) 4 . which is ignited to and weighed as the
dio-xide (.see Section .XI, IIG). Quantitative .separation is thus made from
titanium, iron, vanadium, aluminium, chromium, thorium, cerium, tin, barium,
calcium, copper, bismuth, antimony, and cadmium. If sulphuric acid is

490

GRAVIMETRY XI, 64/65

employed, the concentration should not exceed 5 per cent : higher concentrations
give low results.

Procedure. The solution (20-30 cm^) may contain 0.05-0.2 g Zr, and
should possess a hydrochloric acid content of about 20 per cent.by volume. Add
50 cm^ of 1 6 per cent aqueous mandelic acid solution and dilute to 1 00 cm^ . Raise
the temperature slowly to 85 °C and maintain this temperature, for 20 minutes.
Filter offthe resulting precipitate through a quantitative filter paper, wash it with
a hot solution containing 2 per cent hydrochloric acid and 5 per cent mandelic
acid. Ignite the filter and precipitate to the oxide in the usual manner; a
temperature of 900-1 000 °C is satisfactory. Weigh as Zr 02 .

Note. 1. Bromomandelic acid is a superior reagent for this determination,
but is more expensive. A similar procedure to that above is employed.

Anions

XI, 64. BORATE. Determination of borate as nitron tetrafluoroborate.

Discussion. Boric acid (100-250 mg) in aqueous solution may be deter-
mined by conversion into tetrafluoroboric acid and precipitation of the latter
with a large excess of nitron (see Section XI, IlH) as nitron tetrafluoro-
borate, which is weighed after drying at 1 10 °C. The accuracy is about 1 per cent.

H 3 BO 3 + 4HF ^ HBF 4 + 3 H 2 O

HBF^ + C2oHi6N^ ^ C2oH26N<^-HBF4
Fluoride ion, and weak acids and bases do not interfere, but nitrate, nitrite,
perchlorate, thiocyanate, chromate, chlorate, iodide, and bromide do. Since
analysis of almost all boron-containing compounds requires a preliminary
treatment which ultimately results in an aqueous boric acid sample, this
procedure may be regarded as a gravimetric determination of boron.

Procedure. Place the aqueous sample solution of boric acid (containing
100-250 mg of H3BO3) in a 250-cm^ polythene beaker and dilute to about 60 cm^
with distilled water. Add 15.0cm^ of the nitron solution (1) and 1.0-1,3 gof A.R.
48 per cent hydrofluoric acid (CARE!). Allow the solution to stand for 10-20
hours, and cool in an ice bath for 2 hours. Collect the precipitate in a porcelain
filtering crucible, and wash it with five 10 -cm^ portions of saturated nitron
tetrafluoroborate solution (2) ; drain the precipitate after each washing. Dry at
105-1 10 °C for 2 hours, and weigh as C2oHi6N4-HBF4.

Notes. 1 . Prepare the nitron reagent by dissolving 3.75 g nitron in 25 cm^
5 per cent acetic acid (by volume). Store in a dark bottle.

2. Prepare the wash solution by adding an excess of solid nitron
tetrafluoroborate to 100 cm^ of water and shaking mechanically for 2 hours.

XI, 65. BROMATE AND BROMIDE. Discussion. These anions are both
determined as silver bromide, AgBr, by precipitation with silver nitrate solution in
the presence of dilute nitric acid. With the bromate, initial reduction to the
bromide is achieved by the procedures described for the chlorate (Section XI, 67)
and the iodate (Section XI, 75). Silver bromide is less soluble in water than is the
chloride. The solubility of the former is 0. 1 1 mg dm “ ^ at 2 1 °C as compared with
1.54rngdm“^ for the latter, hence the procedure for the determination of
bromide is practically the same as that for chloride. Protection from light is even
more essential with the bromide than with the chloride because of its greater
sensitivity.

491

XI, 66 QUANTITATIVE INORGANIC ANALYSIS

XI, 66. CARBONATE. Delennination of carbonate by the evolution of carbon
dioxide. Discussion. The carbonate is decomposed by dilute acid, and either
the loss in weight due to the escape of carbon dioxide determined (indirect
method) or the carbon dioxide evolved is absorbed in a suitable medium and the
increase in weight of the absorbent determined (direct method). The direct
method gives more satisfactory results, and will therefore be described. The
indirect method is often employed, however, for samples containing relatively
large amount.s ofcarbonatc.

The decomposition of the carbonate may be cnected with dilute hydrochloric
acid, dilute perchloric acid, or syrupy phosphoric acid. The la.st-named acid is

perhaps the most convenient because of
its comparative non-volatility and the
fact that the reaction can be more easily
controlled than with the other acids. If
dilute hydrochloric acid is employed, a
short, water-cooled condenser should
be inserted between the decomposition
llask and the absorption train (see
below).

Two absorlx-nis are required, one for
water vapour the other for carbon
dioxide. The absorbents for water
vapour which arc generally employed are: (<i) aniiydrous calcium chloride (14-20
mesh), (b) anhydrou.s ctilcium sulphate CDrierile' or ‘Anhydroccl'), and (c)
anhydrous magnesium perchlorate CAnhydrone'). Both (b) and (c) are
preferable to (a); (c) absorbs about 50 per cent of its weight of water, but is
expensive. Anhydrous calcium chloride usually contains ;t little free lime, which
will absorb carbon dioxide ;ilso; it is essential to saturate the U-lube containing
calcium clilonde with dry c;irboa dioxide for several hours and then to disphtce
the carbon dioxide by a current of pure dry air before use.

The absorbents for carbon dioxide in general use are: UD soda lime (this is
available also m the form of self-indicating granules, 'Carbosorb', whicli indicate
when the absorbent is exhausted), (e) .soda lime-asbestos (the 'Carbosorb'
variety gives a marked colour change and therefore indicates the degree of
exhaiLStion), and (J') sodium hydroxide asbe.sios (.'A.scaritc'). In all cases the
carbon dioxide is :ibsorbed in ;iccordance with llie following equation:

2Na01 1 -t- CO, == Na ,COj -f 1 1,0

Water is formed in the re:tetion. lienee it is essenii;il to fill one-quarter or one-
third of the tube with any of the desiccants referred to above (Fig. XI, 2).

Froccihav. Fit up the apparatus shown in Fig. XI. 5. A is a fl:isk of about
100 cm^ capacity, B is a dropping funnel containing 20-2.scm^ of A.R. syrupy
pho.sphoric acid, C is a soda-lime guard tube, D is a bubbler containing syrupy
phosphoric :icid, E is a U-tubc coiilaiiiing calcium chloride which has been
satunitcd with c:irbon dioxide and the residual carbon dioxide displaced by air
(this may be replaced by anhydrous calcium sulphate, or by anhydrous
magnesium perchlorate, ifav:iilable).* F and G are U-lubes containing soda-lime

* rile first ihud of llii'i tube may be fiilcil with anhydrous cupper sulphate to remove any hydrogen
sulphide or hydrogen chloride present from sulphides or chlorides in the limestone.

492

GRAVIMETRY XI, 67

(this absorbent may be replaced by ‘Carbosorb’, soda-lime-asbestos, or
‘Ascarite’), and His a guard U-tube containing the same desiccant as in E. The U-
tubes may be suspended by silver wires attached to hooks on the glass or metal
rod I, or by some other means. All joints are made with short lengths of stout-
walled rubber tubing, and the two ends of the glass tubing should be in contact.
Rubber bungs are employed in A, B, and C. Before proceeding with the actual
determination, make sure that the apparatus is gas-tight.

Weigh out accurately 0.5-0.6g of the carbonate (1) into the flask A, which
should be clean and dry. Remove the two soda-lime or ‘Ascarite’ tubes F and G,
wipe them with a clean linen handkerchief or cloth, and leave them in the balance
case for 45 minutes. Open the taps of the U-tubes momentarily to the air in the
balance case, and weigh them separately. Replace them on the drying train ; place
25 cm^ A.R. syrupy phosphoric acid in B, and see that the apparatus is connected
up as in Fig. XI, 3. Open the taps of the U-tubes. Run in sufficient phosphoric
acid from the tap funnel to cover the solid in the flask (the 25 cm^ will more than
suffice). Close the tap of the funnel and heat the flask carefully; regulate the
temperature so that not more than 2 bubbles of gas per second pass through the
bubbler D. After about. 30-40 minutes, the contents of the flask should be
boiling; boil for 2-3 minutes. Remove the flame, and immediately attach a filter
pump and a large bubbler (similar to D, and containing syrupy phosphoric acid)
to the end of the tube H, Open the tap funnel, and draw air through the apparatus
at the rate of about 2 bubbles per second for 20 minutes. Remove the tubes F and
G, close the taps, treat them as before, and weigh them. From the increase in
weight, calculate the percentage of CO 2 in the sample (2).

Notes. 1 . For practice in this determination, the student may employ A.R.
calcium carbonate or ‘Limestone, 15e’ (Analysed Samples for Students) from the
Bureau of Analysed Samples.

accurate work, and particularly when the arnount of carbon
loxide IS small, a ‘blank’ experiment must be run with the reagents alone before
the determination proper is carried out.

XI, 67. CHLORATE. Determination of chlorate as silver chloride.
iscussioti. ^ The chlorate is reduced to chloride, and the latter is deter-
ined as silver chloride, AgCl. The reduction may be performed with

493

XI, 68/69/70 QUANTITATIVE INORGANIC ANALYSIS

iron(II) sulphate solution, sulphur dioxide, or by zinc powder and acetic acid.
Alkali chlorate.s may be quantitatively converted into chlorides by three
evaporations with concentrated liydrocliloric acid, or by evaporation with thra’
limes the weight ofammonium chloride.

Froccdurc. The chlorate solution should have a volume of about 100 cra^
and contain about 0.2 g CIO^. Add SOcnf* of a 10 per cent solution of A.R.
crystallised iron(ll) sulphate, heat with constant stirring to tlic boiling point, and
boil for 15 minutes. Allow to cool, add nitric acid until the precipitated basic
iron(lll) -salt i.s dissolved, precipitate the chloride by means of silver nitrate
solution, and collect and weigh a.s AgCl after the usual treatment (Section .\I,
17).

Alternatively, treat the chlorate solution with e.xcess of sulphur dioxide, boil
the solution to remove the exce.ss of the gas, render slightly acid with nitric acid,
and precipitate the silver ehloride as above.

For the reduction with zinc, render the chlorate solution strongly acid with
acetic acid, add e.vcess of zinc, and boil the mixture for 1 hour. Dissolve the excess
of unu.scd zinc with nitric acid, filter, and treat the filtrate with silver nitrate in the
usual manner.

Note, Hypochlorites and chlorites may be reduced to chlorides with
sulphur dioxide, and determined in the same way.

XI, 68. CHLORIDE, Piscussioii. This anion is determined as siher
chloride, AgCl; full details arc given in Section XI, 17. Anions which give silver
salts which are insoluble in dilute nitric acid must be absent; these include
bromide, iodide, thiocyanate, sulphide, thiosulphate, hexaeyanoferrate(ll), and
hexacyanoferratedll). Heavy metals interfere, and must be removed by
precipitation.

If the chloride is insoluble, it is necessary to boil it, with a large excess of
saturated sodium carbonate solution or, belter, to fuse it with sodium carbonate
and extract the melt with water. In either case the ehloride pa.sscs into solution,
and is determined in the usual way after acidification with nitric acid.

XI, 69. CY.XNIDE. Di.'iaismin. This anion is determined as .silver cyanide,
AgCN; the experimental details arc similar to those given for Chloride, e.xcepl
that, owing to the volatility of hydrocyanic acid, the solution must not be healed.
The cold solution of alkali cyanide is treated with a slight excess of silver nitrate
solution, faintly acidified w'iih nitric acid, the precipitate allowed to settle,
collected on a weished lillering crucible, and weiehed as AgCN after drying at
100 C.

XI, 71). FLUORIDE. This anion may be determined in one of the follow’ing
forms;

Determination of fluoride as triplienyltin fluoride.*

Discussion. Triplienyltin chloride reagent precipitates fluorides quantitatively as
the eorrespuiuling fluoride. The precipitate is crystalline, easily filtered, and
washed, and is quite stable. Owing to the insolubility of the reagent in water,
precipitation is carried out in 60-70 per cent ethanol solution, and washing is

* Thi;; mclliod Is only of use where .i linilnIegrecofuccut.icy isnot remiiicd.

494

GRAVIMETRY XI, 71

effected with an ethanolic solution of the reagent saturated with triphenyltin
fluoride.

The method is well adapted for the determination of small quantities of
fluorides ; the maximum amount that can be conveniently handled is 0.04 g of F.
The solution should have a pH of 5-7 ; if acid some fluorine will be lost on heating
to boiling, and if basic, triphenyltin hydroxide will be precipitated along with the
fluoride. Metals other than the alkali metals should preferably be absent ; the
latter'may be renroved by washing the precipitate several times with the ethanol
wash solution, followed by cold water. Small quantities of nitrates, chlorides,
bromides, iodides, and sulphates do not interfere, but silicates (with ammoniacal
zinc hydroxide), phosphates (with silver nitrate), and carbonates must be
removed before precipitation.. Carbonate is best removed by neutralising with
dilute nitric acid (to phenolphthaleiii), and boiling off the carbon dioxide. A
disadvantage of the method is that the reagent is expensive.

Procedure. The solution (say, 25 cm^) should contain not more than 0.04 g
of F, and be almost neutral ( 1). Add 95 per cent ethanol to the aqueous solution
of the fluoride so that it comprises about 60-70 per cent of the final volume. Heat
to boiling and treat with twice the calculated quantity of the reagent (2) diluted
with an equal volume of 95 per cent ethanol and also heated to boiling. The latter
is run slowly into the hot fluoride solution with vigorous stirring, and the whole
again heated to the boiling point. Remove the source of heat, and continue the
stirring until the solution has cooled somewhat (3). Allow to stand overnight, and
cool for 1 hour in ice (4). Filter through a weighed, sintered glass or porcelain
filtering crucible, wash with 95 per cent ethanol which has been saturated with
triphenyltin fluoride (about 50 cm^). Dry for 40 minutes at 110°C, cool in a
desiccator, and weigh as (C6H5)3SnF.

Notes. 1. For practice in this estimation, A.R. sodium fluoride may be
used. If desired, a sample of pure sodium fluoride may be prepared as follows.
Treat A.R. anhydrous sodium carbonate with an excess of A.R. hydrofluoric
acid (CARE!) in a platinum dish, and allow to stand for a few hours. Remove the
excess of acid by heating (fume cupboard; hood), allow to cool, and add more
acid. Mix thoroughly with a platinum spatula, heat the dish gently at first, and
then strongly until the sodium fluoride is entirely fused. Pulverise in an agate
mortar, dry the powder in platinum at 1 10 °C, and store in a desiccator over
calcium chloride.

2. The reagent is prepared by shaking vigorously 4.0 g triphenyltin chloride
with 200 cm^ 95 per cent ethanol ; filter from the small undissolved residue. This
is practically a saturated solution.

About 55 cm^ of this reagent are required for 0.04 g F.

3. If the quantity of fluoride is large, precipitation as white . crystals
commences in about a minute after the addition of the reagent, but with small
quantities it does not take place until the solution has cooled to' room
temperature.

4. This is unnecessary if the amount of fluoride is large and the total volume of
the solution small. ■ ,

XI, 7L FLUOROSILICATE. Discussion. The determination of this anion
IS of little practical importance. The methods available for its determination will,
owever, be outlined. Alkali fluorosilicates are decomposed by heating with
so lum carbonate solution into a fluoride and silicic acid :•

495

XI,68/6!)/70 QUANTITATIVE INORGANIC ANALYSIS

iron(n) sulphate solution, sulphur dioxide, or by zinc powder and acetic acid.
Alkali chlorates may be quantitatively converted into chlorides by three
evaporation.s with concentrated hydrocliloric acid, or by evaporation with three
limes the weight orammonium chloride.

Proirdtin-. The chlorate solution should have a volume of about 100cra\
and contain about 0.2 g ClOi. Add iOcin^ of a 10 per cent solution of A.R.
crystallised irondl) sulphate, heat with constant stirring to the boiling point, and
boil for 15 minutes. .Allow to cool, add nitric acid until the precipitated basic
iron(Ul) salt is dissolved, precipitate the chloride by means of silver nitrate
solution, and collect and weigh as AgCl tiftcr tiie usual treatment (Section XI,

Alternatively, treat the chlorate solution with c.xccss of sulphur dio.xide, boil
the .solution to remove the c.xcess of the gas, render slightly acid with nitric acid,
and precipitate the silver chloride as above.

For the reduction with zinc, render the chlorate solution strongly acid with
acetic acid, advl e.xccss of zinc, and hoi! the mixture for 1 hour. Dis.solve the c.xcess
of unused zinc with nitric acid, fdter. and treat the filtrate vviih silver nitrate in the
usual manner.

Note. Hypochlorites .and ehloritvs may be reduced to chlorides with
sulphur dioxide, and determined in the same way.

XI, 6S. CHLORIUEl. Diteuxsitm. This anion is determined as silver
chloride, AgCI ; full details arc given in Section .\T, 17. Anions which give silver
salts which are insoluble in dilute nitric acid must be absent; these include
bromide, iodide, ihiocyamitc. sulphide, thiosulphate, hcxacyanoferratc(ll), and
hc.xacyanoferrate(in). Heavy metals interfere, ami must he removed by
precipitation.

if the chloride is insoluble, it is necessary to boil it, with a large excess of
.saturated sodium carbonate solution or, better, to fuse it with sodium carbonate
and extract the melt with water. In either case the chloride piisses into solution,
and is determined in the usual way after acidilicalion with nitric acid.

.\I, 6!>. CY.AiNlDE. Dhcusimi. This aiiiou is ilcCcrmincd as silver cyanide,
AgCN; the e.\pcrimcntal details are similar to tho.se given for Chloride, c.xcept
that, owing to the volatility of hydrocyanic acid, the solution must not be heated.
The cold solution of alkali cyanide is treated with a slight c.xcess of silver nitrate
solution, faintly aciddied with nunc acid, the precipitate allowed to settle,
collected on a weighed filicrinu crucible, and weiehed as .AeCN after drvingat
100 C. ' " ^

.XI, 70. FLUORIDE, fiiis anion may be determined in one of the following
forms :

Determination of liuoride as tripheiiyliin fluoride.*

Dheussion. T riphenylim chloride reagem precipitates fiuoridc,s quantitatively as
the corresponding iluoride. The prceipiiaie is crystalline, easily filtered, and
washed, and is quite stable Owing to the insolubility of llic reagent in water,
precipitation is carried out in 60-70 per cent ethanol" solution, and washing is

This mclhod is only ol use \\hefc.i high degree of accuracy ivnoi required.

494

GRAVIMETRY. XI, 77

usually determined by precipitation as silver iodide, Agl. Silver iodide is the least
soluble of the silver halides ; 1 dm^ of water dissolves 0.0035 mg at 21 °C. Co-
precipitation and similar errors are more likely to occur with iodide than with the
other halides.

Procedure. Precipitation is therefore made by adding a very dilute
solution, say 0.05M, of silver nitrate slowly and with constant stirring to a dilute
ammoniacal solution of the iodide imtil precipitation is complete, and then
adding excess of nitric acid (1 per cent by volume). The precipitate is collected in
the usual manner, washed with 1 per cent nitric acid, and finally with a little water
to remove nitric acid. Peptisation tends to occur with excess of water. Other
details of the determination will be found in Section XI, 17.

B. Determination of iodide as palladium(ll) iodide. Discussion. Iodide
may also be determined by precipitation as palladium(II) iodide, Pdlj.
Substances, such as ethanol, which cause reduction to metallic palladium must be
absent; bromides and chlorides are not precipitated and therefore do not
interfere. The precipitate is insoluble in water and in dilute hydrochloric acid
(i:99). The reagent, palladium(II) chloride, is expensive, and the method is
therefore rarely employed except for gravimetric separation from other halides.

Procedure. The iodide solution should contain 1 per cent by volume of
hydrochloric acid, and not more than 0.1 g iodide. Warm to 70 °C, and add
palladium(II) chloride solution, dropwise and with stirring, until no more
precipitate is formed. Allow the solution to stand for 24-48 hours at 20-30 °C,
filter the brownish-black precipitate on a weighed filtering crucible (Gooch,
sintered glass, or porcelain), and wash four times with warm water. Dry at
100 °C, for 1 hour, and weigh as Pdij.

XI, 77. NITRATE. Determination of nitrate as nitron nitrate.

Discussion. The mono-acid base nitron, C20H16N4, forms a fairly in-
soluble crystalline nitrate, C2oHi6N4,HN03 (solubility is 0.099 g/dm"^ at
about 20 °C), which can be used for the quantitative determination of nitrates
(see Section XI, HH). The sulphate and acetate are soluble so that precipitation
may be made in sulphuric or acetic acid solution. Perchlorates (0.08 g), iodides
(^•17 g), thiocyanates (0.4 g), chromates (0.6 g), chlorates (1.2g), nitrites (1.9 g),
bronudes (6.1 g), hexacyanoferrate(II), hexacyanoferrate(III), oxalates, and
considerable quantities of chlorides interfere, and should be absent. The figures
in parentheses are the approximate solubilities of the nitron salts in gdm“^ at
about 20 °C.

Procedure. The solution (75-lOOcm^) should be neutral and contain
about 0.1 g NO3. Add 1 cm^ glacial acetic acid or 0.5 cm^ M-sulphuric acid and
heat the solution nearly to the boiling point. Then introduce in one portion
10-1 2 cm^ of the nitron reagent (1), stir, and cool in ice-water for 2 hours. Filter
through a weighed filtering crucible (Gooch, sintered glass, or porcelain). Wash
with 10-1 5 cm^ of a cold saturated solution of nitron nitrate, added in several
portions, and drain the precipitate well after each washing. Finally, wash twice
with 3-cm^ portions of ice-cold water. Dry at 1 05 °C ( I hour is usually required),
and weigh as C2oHi6N4,HN03.

Note. 1. Prepare the reagent by dissolving 5 g of nitron in 50 cm^ of 5 per
cent acetic acid. Store in an amber bottle.

XI, 78, NITRITE, No satisfactory gravimetric procedure is available.
1 itrimetric methods are described in Chapter X.

497

XI, mnn-insiu quantitative inorganic analysis

Na2[SiF6]-l'2N:uCOj + H,0 = 6NaF-rUjSv0j + 2C02

Insoluble fluorosilicates arc brought into solution by fusion svith four times the
bulk' of fusion ini.xturc. and extracting the melt with water. In either case, the
solution is treated with a considerable excess of ammonium carbonate, warmed
to 40 “C, and, after standing for 12 hours, the precipitated silicic acid is filtered
olT, and washed witii 2 per cent ammonium carbonate solution. The filtrate
contains a little silicic acid, which may be removed by shaking with a little freshly
precipitated cadmium oxide. Tlie fluoride in the hltrate is determined as
described in Section XI, 70.

If an acid solution of a tUiorosilicatc is rendered faintly alkaline with aqueous
sodium hydroxide and then shaken with freshly precipitated cadmium o.xide, ail
the silicic acid is' adsorbed by the suspension. The alkali fluoride is then
determined in the filtrate.

.XI, 72. HF„X.ACY.MS'OFKRR,VrE(IIl). No satisfactory gravimetric method
is available. For litrimctric methods, see Chapter X.

XI, 73. HFXACYANOFERKA'rE(H). No satisfactory gravimetric pro-
cedure is available. Titriineiric methods arc described in Cliapter X.

XI, 74. HYPOFflCSFinTTv, This anion is determined similarly to
phosphite (Section XI, 82) either indirectly a.s lucrcurj(I) chloride, HgiClj, eras
ammonium inaj>nesium phosphate he.vahydrate, Mg.NH^FOi.filljO, or as
magnesium pyrophosphate, .MgjICO,. In this case the reaction with mercur)(ll)
chloritlc solution is:

4MgCl, + ll,PO.-!-2H,a ^ 2Hg>C!>'rH,PO.iT4HC!
so that 2HgX'l. :: HiPO.

.XI, 75. lO D.VTF, Determination of iodale as sib er iodide.

Disaission. lodates are readily reduced by sulphurous acid to iodides; the
latter arc determined by precipitation with silver nitrate solution as siher
iodide, Agl. lodates cannot be converted quantitatively into iodides by ignition,
lor the decomposition takes place .it a temperature at which the iodide is
appreciably volatile.

Periodates are also reduced by sulphurous acid, and may therefore be similarly
determined. Similar remarks apply to broniales; these arc ultimately weighed as
silver bromide. AgBr.

Procedure. Acidify tlie iodatc .solution (lOOcnr' cotilainirig co. 0.3g of
lOj) (1) vvith suipluinc acid, and pass in .sulphur dioxide (or add a freshly
prepared saturated solution of sulphurous acid) until tlie solution which at first
becomes yellow, on account ot tlie separation of iodine, i.s again colourless. Boil
oir the excess of sulphur dioxide, and precipitate the iodide with dilute silver
nitrate solution as de.scribed in .Section XI, 76. Weigh as Agl.

Note. 1. For practice in this determination, A. R. potassium iodate may be
employed.

XI, 76. IODIDE. Two procedures are commonly employed for the
determination of iodides.

A. Determination of iodide as silver iodide. Discussion. This anion is

496

GRAVIMETRY XI, 77

usually determined by precipitation as sOver iodide, Agl. Silver iodide is the least,
soluble of the silyer halides; 1 dm^ of water dissolves 0.0035 mg at 21 °C. Co-
precipitation and similar errors are more likely to occur with iodide than with the
other halides. . . '

Procedure. Precipitation is therefore made by adding, a very dilute
solution, say 0.05iW, of silver nitrate slowly and with constant stirring to a dilute
ammoniacal solution of the iodide until precipitation is complete, and then
adding excess of nitric acid ( 1 per cent by volume). The precipitate is collected in
the usual manner, washed with 1 per cent nitric acid, and finally with a little water
to remove nitric acid. Peptisation tends to occur with excess of water. Other
details of the determination will be found in Section XI, 17.

B. Determination of iodide as palladium(n) iodide. Discussion. Iodide
may also be determined by precipitation as palladium(II) iodide, Pdlj.
Substances, such as ethanol, which cause reduction to metalhc palladium must be
absent; bromides and chlorides are not precipitated and therefore do not
interfere. The precipitate is insoluble in water and in dilute hydrochloric acid
(1:99). The reagent, palladium(II) chloride, is expensive, and the method is
therefore rarely employed except for gravimetric separation from other halides.

Procedure. The iodide solution should contain 1 per cent by volume of
hydrochloric acid, and not more than 0.1 g iodide. Warm to 70 °C, and add
pal]adium(II) chloride solution, dropwise and with stirring, until no more
precipitate is formed. Allow the solution to stand for 24-48 hours at 20-30 "C,
filter the brownish-black precipitate on a weighed filtering crucible (Gooch,
sintered glass, or porcelain), and wash four times with warm water. Dry at
100 °C, for 1 hour, and weigh as Pdij.

XI, 77. NITRATE. Determination of nitrate as nitron nitrate.

Discussion. The mono-acid base nitron, C20H16N4, forms a fairly in-
soluble crystalline nitrate, C2oHi6N4,HN03 (solubility is 0.099 g/dm"^ at
about 20 °C), which can be used for the quantitative determination of nitrates
(see Section XI, IIH). The sulphate and acetate are soluble so that precipitation
may be made in sulphuric or acetic acid solution. Perchlorates (0.08 g), iodides
(0.17g), thiocyanates (0.4 g), chromates (0.6 g), chlorates (1.2g), nitrites (1.9 g),
bromides (6. 1 g), hexacyanoferrate(II), hexacyanoferrate(III), oxalates, and
considerable quantities of chlorides interfere, and should be absent. The figures
m parentheses are the approximate solubilities of the nitron salts in gdm“^ at
about 20 °C.

Procedure. The solution (75-100 cm^) should be neutral arid contain
about 0.1 g NO3. Add 1 cm^ glacial acetic acid or 0.5 cm^ M-sulphuric acid and
heat the solution nearly to the boiling point. Then introduce in one portion
10-I2cm^ of the nitron reagent (1), stir, and cool in ice-water for 2 hours. Filter
through a weighed filtering crucible (Gooch, sintered glass, or porcelain). Wash
with 10-15 cm^ of a cold saturated solution of nitron nitrate, added in several
portions, and drain the precipitate well after each washing. Finally, wash twice
with 3-cm^ portions of ice-cold water. Dry at 105 °C (1 hour is usually required),
and weigh as C2oH,6N4,HN03.

Note. 1. Prepare the reagent by dissolving 5 g of nitron in 50 cm^ of 5 per
cent acetic acid. Store in an amber bottle.

XI, 78. NITRITE. No satisfactory gravimetric procedure is available.
1 itrimetric methods are described in Chapter X.

497

XI, 79/S0/81 QUANTITATIVE INORGANIC ANALYSIS

XI, 79. OX.ALATtL Determination of o.valate as calcium oxalate and as
calcium carbonate or calcium oxide. DLscu.sxioii. The neutral solution of alkali
oxalate is acidified with acetic acid, healed to boiling, and precipitated with
boiling calcium chloride solution. After standing for 12 hours, the precipitate is
filtered off, washed w'ith hot water, and weighed cither as calcium oxalate, or
after heating, as calcium carbonate. CaCO,, or a.s calcium oxide, CaO.

Proccduri’, The following rapid method yields results of moderate
ticcuracy. Precipitation of the oxalate is effected in boiling solution containing a
little ammonium chloride by a hot solution ofcalcium chloride. The solution is
allowed to cool, treated with one-third of its volume of 90 per cent ethanol, and
allowed to stand for .lO minutes. The precipitate is washed by decantation
through a weighed porcelain or sintered glass filtering crucible with warm water
(50-60 C) until the chloride reaction is negative. The calcium oxalate is then
transferred to tlie filtering crucible, washed once w itlr cold water, five times with
ethanol, and several tinie.s w ith small volumes of anhydrous diethyl ether. The
precipitate is sucked dry at the pump for 10 minutes, the outside of ihe crucible
wiped dry with a clean linen cloth, and then left in a vacuum desiccator for 10
minutes. It is weighed as CaC, 04 ,H, 0 , or may be converted to the other two
forms (see Section ,\1, 20).

XI, 80. PEIRCIiLOR.AT'Il Determination of perchlorate as silver chloride.
Di'sciisxion. Perchlorates are not redticetl by iron(Il) sulphate solution,
sulphurous acid, or by repeated evaporation with concentrated hydro-
chloric acid; reduction occurs, however, with titanium(Hi) sulphate
solution. Ignition of perchlorates with ammoniimt chloride in a platinum
crucible or in a porcelain crucible in the presence of a little platinum powder
results in reduction to the clilorides (the platinum acts as a catalyst), which may
be determined in the usual manner. Losses occur when perchlorates arc ignited
alone.

Proci’durc. The perchlorate, if .supplied as a solution, is evaporated to
diyness on the water bath; otlierwisc the solid fTerchlorate is used directly,
intimately mix about O.-lg of the perchlorate (!) with 1.5g of A.R. ammonium
chloride in a pluiiiium crucible covered with a watch ghiss or lid. ignite gently
until fuming ceases and coittinue the heating for 1 hour. Do not fu-se tlie resulting
chloride, as the crucible may be attacked. Repeal the ignition witli another 1.5 g
of ammonium clilonde. Dissolve the residue in a little water, filter through a
small ciuamittitive filter paper to remove any platinum powder which may be
present, and determine the chloriiic in the filtrate as silver chloride (Section XI,
17).

Note. 1. For practice in this determination, employ A.R. potassium
perchlorate.

.XI, 81. RHOSPHATE. Phosphates may be determined by cither of the
following methods.

A. Dctcrmiimtion of pho.spliatc as ammonium magnesium phosphate
liexahydrate or as magne.sium pyropho.sphate. Dixaixsio)). Orlhopliosphates
may be precipitated as ammonium magnesium phosphate, MgNH 4 P 04 , 6 H, 0 , by
magnesium chloride and ammonium chloride in aminoniacal solution

498

GRAVIMETRY XI, 81

(‘magnesia’ reagent). Most elements, other than those of the alkalis, interfere,
however, by giving precipitates with ‘magnesia mixture’.* It is therefore
necessary in the majority of cases to separate the phosphate first from interfering
substances. This may be readily effected by precipitation as ammonium
molybdophosphate with excess of ammonium molybdate in warm nitric acid
solution; arsenic, vanadium, titanium, zirconium, silica, and excessive amounts
of ammonium salts interfere. When -first precipitated (in the presence of a large
excess of nitric acid and of ammonium nitrate), the yellow precipitate has the
composition (NH4)2H[PMoi204o],H20. Upon washing with a dilute solution
of ammonium nitrate, the diammonium salt passes easily into the triammonium
salt (NH4)3 [PMoi 204 o]- The precipitate thus obtained is dissolved in dilute
ammonia solution, and the phosphate is then precipitated as ammonium
magnesium phosphate. A double precipitation of the latter is usually necessary in
order to obtain a precipitate entirely free from molybdate.

Procedure. To a neutral or weakly acid solution (50-100 cm^) of the
phosphate, containing not more than 0.10 g of P2O5 and free from interfering
elements (1), add 3cm^ of concentrated hydrochloric acid and a few drops of
methyl red indicator. Introduce 25 cm^ of magnesia mixture (2), followed by pure
concentrated ammonia solution slowly, whilst stirring the solution vigorously
until the indicator turns yellow. The procedure from this stage is the same as
described for the determination of magnesium in Section XI, 23, except that
when carrying out the re-precipitation from the hydrochloric acid solution 2 cm^
of the magnesia mixture are added instead of the 1 cm^ of ammonium phosphate
solution. Weigh as MgNH4P04,6H20 or as Mg2P207.

Note. 1. A suitable solution for practice may be prepared by dissolving
about 0.4g, accurately weighed, A.R. anhydrous Na2HP04 in 100 cm^ water.
The appropriate weight of A.R. KH2PO4 may also be used and is perhaps to be
preferred.

2. The magnesia mixture is prepared as follows. Dissolve 25 g magnesium
chloride MgCl2,6H20 and 50 g ammonium chloride in 250 cm^ of water. Add a
slight excess of ammonia solution, allow to stand overnight, and filter if a
precipitate is present. Acidify with dilute hydrochloric acid, add Icm^
concentrated hydrochloric acid, and dilute to 500 cm^.

B. Determination of phosphate as ammonium molybdophosphate.

Discussion. If interfering elements are absent, the original yellow precipitate
obtained in A. above may be weighed either as ammonium molybdophosphate,
(^4)3 [PMoj 204 o], after drying at 200-400‘’C (280 °C is recommended) or
as P20 s, 24 Mo 03, after heating at 800-825 °C for about 30 minutes. For practice
m this determination the student may determine the percentage of P2O5 (or P) in
anhydrous A.R. disodium hydrogen phosphate Na2HP04 or, preferably, in
A.R. potassium dihydrogen phosphate KH2PO4. Some experimental details will
e found in Section X, 49, a slightly modified procedure is described below.

Procedure. Prepare a solution of anhydrous A.R. disodium hydrogen
or of A.R. potassium dihydrogen phosphate containing about 125 mg
0 °205in 150cm^. Warmto60°C,andrunin 100 cm^ of ammonium molybdate

Phosphate may be precipitated directly as ammonium magnesium phosphate in the presence of
ements such as iron, aluminium, titanium, zirconium, tin, and calcium by adding excess of citric
acia and usmg an excess of magnesia mixture.

499

xr, 82 QUANTJTATl VE INORGANIC ANA LYSIS

reagent (1) also warmed to 60 C: use a fast-flowing pipette for the addition and
stir well. Heat to 60 ’C for about 1 hour with frequent stirring. Collect the
precipitate in a weighed porcelain filtering crucible using two 20-cm^ portions of
2 per cent ammonium nitrate solution to transfer it from the beaker (remove any
precipitate adhering to the walls of the beaker with a rubber-tipped glass rod);
wash the precipitate in the crucible with five lO-cm-^ portions of 2 per cent
ammonium nitrate solution. Dry the precipitate at 280 'C, and weigh :is
(NH^ljtPMOijO^ol- an additional check, ignite the precipitate at 800-
825 'C in an electric mulile furnace; weigh as P,05.24MoOj. Both solids arc
appreciably hygroscopic; the covered crucible, after cooling in a desiccator,
should be weighed as soon as it has acquired the laboratory' temperature.

Note. 1. Prepare the aninionium molybdate reagent as follows. Dissolve
125g ammonium nitr.iie in 125em'’ water in a fiask and add 175cm^ nitric acid,
sp.gr. 1.42. Dissolve 12.5 g A. R, ammonium molybdate* in 75 cm^ of water and
add this slowly and with constant shaking to the nitrate solution. Dilute to
500 enr’ with water, heat the flask in a water bath at 60 'C for 6 hours, and allow
the solution to stand for 24 hours. If a precipitate forms, filter through a
Whatman No. 42 filter paper. This reagent has good keeping qualtlies; it is said
that no precipitate is formed for at le,j.si .4 months.

XI, 82. PHOSPHITE This anion may be determined in either of the
following forms:

A. Determination of phosphite as mercury (1) chloride. Dhaasioit. The
acid solution of phosphite reduces mercuty(ll) chloride solution to mercuryd)
chloride which is weighed. The reaction is:

2HgCU + HjPO, + H>0 - He,Cl. i-H 3 P 04 -r 2 HCl

whence Hg.Clj H^PCj

Procedure. The phospliitc solution (.40cm'’) should contain about O.lg
HPOj*'. Place 50cm'’ of .4 per cent mcrciirydl) chloride solution, 20cm’ of 10
percent sodium acetate, and 5 cm’ of glacial acetic add in a 250-cm’ beaker, and
'.sdd the phospiiile solution dropwisc, and with .stirring, in the cold. Allow to
stand on a water bath at 40 50 C fiir 2 hours. Wiicn cold, filler through a
weighed filtering crucible (Gooch, sintered glass, or procekiin). wash two or three
time.s with 1 per cent hydrochloric acid, and then four times with warm water.
Dry at 105-1 10 C, and weigh as HgXN.

B. Determination of plursphite a.s anmiunium magnesium phosphate
he.xahydrate or a.s the pyropho-spliale. Di<.cm.sion. The phosphite is o.vidised by
nitric acid to phosphate, and the latter is determined us ammonium magnesium
phosphate hexahydraie or as the pyrophosphate.

Procedure. Treat the aqueous .sulultoiv of the phosphite (lOOcm’) with
5cm’ conecniraled nitric acid, evaporate to a small volume on the water bath,
add 1 cm’ fuming nitric acid, and heat again. Dilme the solution, and precipitate
the phosphoric acid by magnesia mixture and ammonia solution, and weigh as
MgNH 4 P 04 , 6 lljO or us Mg,P,D, (Section XI, 81.V).

* 4hisisaclu.illy ihchcpU-mol)bU;ile(Nllj)„Mo.O,4.ailjO.

500

GRAVIMETRY XI, 83

XI, 83. SILICATE. For analytical purposes silicates may be conveniently
divided into the following two classes: (a) those (‘soluble’ silicates) which are
decomposed by acids, such as hydrochloric acid, to form silicic acid and the salts
(e.g., chlorides) of the metals present, and (b) those (‘insoluble’ silicates) which
are not decomposed by any acid, except hydrofluoric acid. There are also many
silicates which are partially decomposed by acids ; for our purpose these will be
included in class (Z)).

A. Determination of sUica in a ‘soluble’ sUicate. Discussion. Most of the
silicates' which come within the classification of ‘soluble’ silicates are the
orthosilicates formed from Si 04 ‘^~ units in combination with just one or two
cations. More highly condensed silicate structures give rise to the ‘insoluble’
silicates.

Procedure. Weigh out accurately about 0.4 g of the finely powdered
silicate (1) into a platinum or porcelain dish, add 10-15 cm^ water, and stir until
the silicate is thoroughly wet. Place the dish, covered with a clock glass, on the
water bath, and add gradually 25 cm^ 1 1 1 hydrochloric acid. The contents of the
dish must be continuously stirred with a glass rod ; when no gritty particles
remain, the powder will have been completely decomposed. Evaporate the
liquid to dryness: stir the residue continuously and break up any lumps with
the glass rod. When the powder appears to be dry, place the basin in an air oven at
100-1 10 °C for 1 hour in order to dehydrate the silica. Moisten the residue with
5 cm^ of concentrated hydrochloric acid, and bring the acid into contact with the
solid with the aid of a stirring rod. Add 75 cm^ of water, rinse down the sides of
the dish, and heat on a steam bath for 10-20 minutes to assist in the solution of
the soluble salts. Filter off the separated silica on a Whatman No. 41 or 541 filter
paper. Wash the precipitate first with warm, dilute hydrochloric acid (approx.
0.5A/), and then with hot water until free from chlorides. Pour the filtrate and
washings into the original dish, evaporate to dryness on the steam bath, and heat
in an air oven at 100-1 10 °C for 1 hour. Moisten the residue with 5cm^
concentrated hydrochloric acid, add 75 cm^ water, warm to extract soluble salts,
and filter through a fresh, but smaller, filter paper. Wash with warm dilute
hydrochloric acid (approx. 0. 1 M), and finally with a little hot water. Fold up the
moist filters, and place them in a weighed platinum crucible. Dry the paper with a
small flame, char the paper, and burn off the carbon over a low flame ; take care
that none of the fine powder is blown away. When all the carbon has been
oxidised, cover the crucible, and heat for an hour at the full temperature of a
Meker type burner in order to complete the dehydration. Allow to cool in a
desiccator, and weigh. Repeat the ignition, etc., until the weight is constant.

To determine the exact SiOj content of the residue, moisten it with 1 cm^
water, add 2 or 3 drops concentrated sulphuric acid and about 5 cm^ of the purest
available (A.R.) hydrofluoric acid. (CARE!) Place the crucible in an air bath
(Section ID, 25) and evaporate the hydrofluoric acid in a fume cupboard (hood)
with a small flame until the acid is completely expelled; the liquid should not be
oiled. (The crucible may also be directly heated with a small non-luminous
ame.) Then increase the heat to volatilise the sulphuric acid, and finally heat
with a Meker-type burner for 15 minutes. Allow to cool in a desiccator and weigh,
^^e-heat to constant weight. The loss in weight represents the weight of the silica

Notes. 1. For practice in this determination, powdered, fused sodium
silicate may be used.

501

XI, 83 QUANTITATIVE INORGANIC ANALYSIS

2. It is advisable; to carry out a blank clcicnnination with the hyUrofluoricacid,
and to allow for any non-volatile substances, if necessary.

B. DeterniinaUon of silica in an “iasoluble’ .silicate, and ultimate neighing as
silica, SiOj. Discussion. Insoluble .silicates are generally fused with sodium
carbonate, and the ntcit, which contains the silicate in acid-decomposable form,
is then treated with hydrochloric acid. The acid solution of the decoraposal
.silicate is evaporated to dryness on the water bath to separate the gelatinous
silicic acid SiOv.-clLO as insoluble silica SiO,.yH,0: the residue is heated at
1 1 0 -1 20 “C to partially dehydrate the silica and render it as insoluble as possible.
The residue is c.xtracted with hot dilute hydrochloric acid to remove salts of iron,
alurainiuin. and other metal.s which may be present. The greater portion of the
silica reniain.s undi.s,solved. and is filtered ulT The filtrate is evaporated to
dryness, and the residue heated at n0“120‘C as before in order to render
insoluble tlie small amount of silicic acid that has escaped dehydration. The
residue is treated with dilute hydrochloric acid as before, and the second portion
of silica is filtered olfon a fresit filler. Tlte two washed precipitates are combined,
and ignited in a platimim crucible at about 1050 C to .silicon dioxide, SiOj, and
the latter is weighed. The ignited residue i.s not usually pure silicon dioxide; it will
generally contain smalt amounts of the o.xides of iron, aluminium, titanium, etc.
The amount of impurity may be determined, if de.'>ired, by treating the weighed
residue in the platinum crucible with an e.\ces.s of hydroiluoric acid and a little
concentrated sulphuric acid. The silica is c.xpcilcd as the volatile silicon
letratluoride; the impurilics (e.g., .Al.Oj; and FctO^) are first converted into the
fluorides, which pass into the sulphates in contact with the less-volatile sulphuric
acid, wliilsi the subsequent brief ignition (at 105U-1 100 'C for a few minutes)
converts the sulphates back into oxides. Thus, for example;

SiO.,-f 6 HF lI,(Sit-\.l + 2H,0
HJSiFJ = .SiF. i-2HF
AKOi + 6 HF = 2 AIF 3 f .1H,0

2 AiFji + 3H,SO.s Ai.iSO^), i-6HF

AI,(S0J3 = AJ,Oi-f3SOi

The loss in weight therefore represents the aiiunint of pure silicon dioxide
present,

Procdcltirc. Weigh out accurately into a platimim crucible about l.Og of
the finely powdered dry silicate (1 ), add six times the weight of anhydrous A.R.
•sodium carbonate (or, better, of A.R. fusion mi.xturc). and mix the solids
thoroughly by stirring with .1 thin, rounded ghuss rod. Cover the mixture with a
little more of the carbon.ite, and then cover the crucible. Heat the mixture
gradually until after about 20 minutes a tranquil melt is obtained ; the cover of the
crucible is lifted occasionally to e.xmninc the contents. Maintain the temperature
of a quiet liquid fusion for about 30 minutes. Allow to cool. Place the crucible
and lid in a covered deep porcelain or platinum basin (or in a large casserole),
cover it vvitli water, and leave overnight, or warm on the water bath until the
contents are well disintegrated. Introduce very slowly by means of a pipette or a
bent funnel about 25cnr' of concentrated hydrochloric acid into the covered
vessel. Warm on the steam bath unii! the evolution of carbon dioxide has ceased.
Remove and rinse the cover glass, crucible, and lid, and evaporate the contents of

502

GRAVIMETRY XI, 83

the dish to complete, dryness on' the steam bath, crushing all lumps with, a glass
rod. Heat the residue for an hour at 100-1 10°C to dehydrate the silica. Complete
the determination as described in A.

Note. ,1. ‘Feldspar (Potash), No. 29dG’ (one of the Analysed Samples for
Students) available from the Bureau of Analysed Samples is suitable.

C. Determination of sUica. in an ‘insoluble’ sUicate as quinoline
molybdosilicate. Discussion. Silica may also be determined gravimetrically as
quinoline molybdosilicate. The solution of silicic acid is treated with ammonium
molybdate to form molybdosilicic acid H4[Si04',12Mo03], which is then
precipitated as quinohne molybdosilicate, (C9H7)4H4[Si04,12Mo03]. The
latter is weighed after drying at 1 50 °C. The experimental conditions lead to
quinoline molybdosilicate in a pure form suitable for weighing.

Phosphate, arsenate, and vanadate interfere. Borate, fluoride, and large
amounts of aluminium, calcium, magnesium, and the alkali metals have no effect
in the determination, but large amounts of iron ( > 5 per cent) appear to produce
slightly low results.

Procedure. The method to be described is especially suitable for ceramic
materials such as fireclay, firebrick, or silica brick. The finely ground sample
should be dried at 1 10 ‘'C. The weight of sample to be employed depends largely
upon the silica content of the material, since not more than 35-40 mg of silica

should be present in the aliquot employed for the determination. For samples
containing more than 65 per cent SiOj ( 1 ), use 0.25 g ; for samples containing less
than 65 per cent Si02, use 0.50 g (2).

Place 7 g of A.R. sodium hydroxide pellets in a nickel crucible (4.5 x 4.5 cm)
and fuse gently until the water is expelled and a clear melt results. Allow to cool,
introduce the weighed sample evenly on to the solidified melt, moisten with a
little ethanol and gently evaporate the ethanol on a hot plate; this reduces the
tendency to spirting in the subsequent fusion. Heat gently over a Bunsen burner,
with occasional rotation of the crucible, until the sodium hydroxide is just

molten, after which raise the temperature to a dull red heat for 2-5 minutes ; the
sample should then have dissolved completely. Carefully cool the crucible by
partial immersion in cold water ; when the melt has just solidified transfer the hot
crucible to a 400-cm^ nickel beaker and cover with a clock glass. Raise the clock
glass slightly, fill the crucible with boiling water and replace the cover ; this should
suffice to dissolve the fused mass, otherwise add a little more boiling water. When
the vigorous reaetion has subsided, wash the clock glass and sides of the beaker
with hot water ; remove the crucible with clean tongs, carefully rinsing it inside
and out with hot water. Dilute the suspension to 175cm^; do not exceed .this
volume. Place 20 cm^ of concentrated hydrochloric acid in a 500-cm^ conical
flask; pour the fusion extract, with swirling, into the acid, rinse the beaker with a
ittle hot water, and add the rinsings to the flask. Cool rapidly to room
and dilute to 250 cm^ in a graduated flask.

aliquot part containing about 35 mg of silica, dilute it to about
0 cm in a 800-cm^ beaker, add 3 g of sodium hydroxide pellets and swirl until
issolved. Add 10 drops of thymol blue indicator (0.04 per cent solution in dilute
e anol, 1.4) followed by concentrated hydrochloric acid dropwise, swirling
constantly, until the colour of the indicator changes from blue, through yellow,
fTi^ solution to become too hot (3). Now add 10 cm,^ of

I u e hydrochloric acid (1 19) and dilute to 400 cm^. Add 50 cm^ of 10 per.cent
nimomum molybdate solution (4) from a burette; stir vigorously during the

503

XI, 84 QUANTITATIVE INORGANIC ANALYSIS

addition and for 1 minute aftorwards. Allow to stand lor !0 minutes, add SOcnt^
of concentrated hydrocliloiic acid, and ininicdiatcly precipitate the yellow
molybdosilicate by introducing 50cm-* of the quinoline reagent (5) from a
burette, stirring constantly. A cream-coloured, finely-divided precipitate of
quinoline molylxiosilicate forms. Warnt the suspension to about 80“C during
about 10 minutes and maintain this temperature for 5 minutes in order to
coagulate the precipitate. Cool in nmning water below 20 C and collect the
precipitate in a .sintered glass filtering crucible fporosily No. 4); wash the
precipitate si.\ times with the special wash solution ( 6 ), taking care not to allow
the precipitate to run dry during the filtration and washing. Dry at 1 50 C for 2
hours and cool the covered crucible in a desiccator. Weieh as
H4SiO.;.)2.MoO,,l.

Notes. 1. ‘Silica brick. No. 267' (a British Chemical Standard) may be
used.

2. 'Firebrick. No. 269' (a British Chemical Standard) may be used.

.1. This prtK-e.ss is to ensure that the silica is in the correct form for reaction
with ammonium molybdate. If the .solution is too hot, the red colour may not
develop.

4. Prepare the 10 per cent atnmoniuni molybdate solution by dissolving 25 g
ammonium molybdate in water and diluting to 25Ucm* in a polythene bottle. It
keeps for about 4 weeks.

5. Prepare the 2 per cent quinoline hydrochloride solution by adding lOcra*
pure quinoline to about 4U0enT' hot water containing 12.5em-* of concentrated
hydrochloric acid, and stirring constantly. Cool the solution, add a little ashless
filler pulp, and leave to .settle. Filter the .solution through a paper pulp pad, but
do not wash. Dilute the filtrate with water to jtlOeuf*.

6 . Prepare the wash solution by diluting 5ctn-* of the 2 per cent quinoline
hydrochloride solution with water to 200 cm'',

XI, 84. SULPHATE. Determinaiion of sulphate a.s barium sulphate.
DiscussUm. The method con,sists in slowly adding a dilute .solution of barium
chloride to a hot solution of the sulpliaie slightly acidified with hydrochloric -acid;

Ua-' +SO 4 -' -- BaSO^

The precipitate is (iliered olf. washed with water, carefully ignited at a red heat,
and weighed as barium sulphate.

The reaction upon wliich the determination depends appears to be a simple
one. but is in reality subject to numerous possible errors: sali.sfuctory rc.sults can
be obtained only if the c.xpcrimental conditions are carefully controlled. Before
some of these are discu,ssed, the student is lecummeitdcd to read Sections XI,3--6.

Barium sulphate has a solubility in water of about 3 mg dm ' ■* at the ordinary
temperature. The solubility is increased in the presence of mineral acids, because
of the formation of the hydrogensulphatc ion (S 04 ' “ + IF ^ HSO 4 ' I: thus the
solubilitie.s at room icinperaliire in the presence of 0.1, 0.5, 1.0, and 3.0.U-
hydrochloric acid are 10.47.87,and 101 mg dm '■* respectively, but the solubility
is less in ilie presence of :i moderate c.sce.ss of bariimi ions. Nevertheless, it is
customary to carry out the precipitation in weakly acid solution in order to
prevent the possible formation of the barium salts of such anions as chromate,
carbonate, and pliosphate, which are insoluble in neutral solutions; moreover,

504

GRAVIMETRY XI, 84

the precipitate thus obtained consists of large crystals, and is therefore more
easily filtered (compare Section XI, 4). It is also of great importance to carry out
the precipitation at boiling temperature, for the relative supersaturation is less at
higher temperatures (compare Section XI, 4). The concentration of hydrochloric
acid is, of course, limited by the solubility of the barium sulphate, but it has been
found that a concentration of O.OSMis suitable; the solubility of the precipitate
in the presence of barium chloride at this acidity is negligible. The precipitate may
be washed with cold water, and losses, owing to solubility influences, may be
neglected except for the most accurate work.

Barium sulphate exhibits a remarkable tendency to carry down other salts (see
co-precipitation. Section XI, 5). Whether the results will be low or high will
depend upon the nature of the co-precipitated salt. Thus barium chloride and
barium nitrate are readily co-precipitated. These salts will be an addition to the
true weight of the barium sulphate, hence the results will be high, since the
chloride is unchanged upon ignition and the nitrate will yield barium oxide. The
error due to the chloride will be considerably reduced by the very slow addition of
hot dilute barium chloride solution to the hot sulphate solution, which is
constantly stirred ; that due to the nitrate cannot be avoided, and hence nitrate
ion must always be removed by evaporation with a large excess of hydrochloric
acid before precipitation. Chlorate has a similar effect to nitrate, and is similarly
removed.

In the presence of certain cations (sodium, potassium, lithium, calcium,
aluminium, chromium, and iron(III)), co-precipitation of the sulphates of these
metals occurs, and the results will accordingly be low. This error cannot be
entirely avoided except by the removal of the interfering ions. Aluminium,
chromium, and iron may be removed by precipitation, and the influence of the
other ions, if present, is reduced by considerably diluting the solution and by
digesting the precipitate (Section XI, 5). It must be pointed out that the general
method of re-precipitation, in order to obtain a purer precipitate, cannot be
employed, because no simple solvent (other than concentrated sulphuric acid) is
available in which the precipitate may be easily dissolved.

Positively charged barium sulphate, which is obtained when sulphate is
precipitated by excess of barium ions, can be coagulated by the addition of a trace
of agar-agar. About 1 mg of agar-agar as a 1 per cent aqueous solution will cause
the flocculation of about 0. 1 g of barium sulphate, but in practice somewhat
larger quantities are generally used. The resulting precipitate does not creep up
the sides of the vessel.

Negatively charged barium sulphate, obtained in the determination of barium
appreciably improved by agar-agar ; this precipitate as a rule, presents little
difficulty in filtration.

Pure barium sulphate is not decomposed when heated in dry air until a
temperature of about 1400 °C is reached :

BaSO.^ =BaO + S 03

however, easily reduced to sulphide at temperatures above
C by the carbon of the filter paper :

BaS 04 -)- 4 C = BaS-f4CO

The reduction is avoided by first charring the paper without inflaming, and then

505

XT, 84 QUANTITATIVE INORGANIC ANALYSIS

addition and for 1 minute afterwards. Allow to stand for 10 minutes, add SOcra^
of concentrated hydrochloric acid, and immediately precipitate the yellow
molybdo-silicate by introducing SOcm-* of the quinoline reagent (5) from a
burette, stirring constantly. A cream-coloured, finely-divided precipitate of
quinoline molybdosilicate forms. Warm the su.spcnsion to about 80 ‘"C during
about 10 minutes and maintain this temperature for 5 minutes in order to
coagulate tlie precipitate. Cool in running water below 20X and collect the
precipitate in a sintered glass lillering crucible (porosity No. 4); wash the
precipitate six times with the special wash .solution (6), taking care not to allow
the precipitate to run dry during the filtration and washing. Dry at 1,50 X for 2
hours and cool the covered crucible in a desiccator. Weigh as
(C,H,)4H^ISiO,.12.MoO,l.

Notes. 1. ‘Silica brick. No. 267’ (a British Chemical Standardi maybe
used.

2. ‘Fiicbrick, No. 269’ (a Uriiisii Chemical Standard) may be used.

3. This proccxs is to ensure that the silica is in the correct form for reaction
with ammonium molybdate. If the solution is too hot. the red colour may not
develop.

4. I’repare the 10 per cent ammonium molybdate solution by dissolving 25 g
ammonium molybdate in water and diluting to 250cin-* in a polythene bottle. It
keeps for about 4 weeks.

5. Prepare ihe 2 per cent quinoline hydrochloride solution by tidding lOcrn^
pure quinoline to about 400cm'' hot water containing 1 2,5cm'' of concentrated
hydrochloric acid, and stirring constantly. Cool the solution, add a little ashlcss
filler pulp, and lease to <eltlc. Filter the solution through a paper pulp pad, but
do not wash. Dilute the filtrate with svaicrto5U0cm'.

6. Prepare the wash solution by diluting 5 cm'' of the 2 per cent quinoline
hydrochloride .solution svith water to 200cm''.

.XI, 84. SULPITA'ITC Determination of sulphate as barium sulphate.
Discu.ssii)ii. The method consists in slowly adding a dilute .solution of barium
chloride to a hot solution of the sulphate slightly acidified with hydrochloric acid:

Ba-'+SOr' - BaS 04

The precipitate is lillered oil', washed with water, carefully ignited at a red heat,
and weighed a-s barium sulphate.

The reaction upon wliich the determination depends appears to be a simple
one, but is in reality subject to numerous possible errors; satisfactory results can
be obtained only if the e.vperiniental conditions arc carefully controlled. Before
some of these arc discu.ssed, the student is recommended to read Sections XI, 3-6,

Barium sulphate has a solubility in water of about 3 mg dm"' at the ordinary
temperature. The solubility i.s increased in the presence of mineral acids, because
of the formation of the hydrogcnsulpliaie ion (SO^' " -r H * c^iHSO^'); thus the
solubilities at room temperature in the presence of 0.1, 0,5, 1.0, and 2.0.1/-
hydrocliloricacid are 10, 47, 87, and 101 mg dm'"' respectively, but the solubility
is less in the presence of a mciderale c.vcoss of barium ions. Nevertheless, it is
customary to carry out the precipitation in weakly acid solution in order to
prevent the possible formation of the barium salts of such anions as chromate,
carbonate, and phosphate, which are insoluble in neutral solutions; moreover.

504

GRAVIMETRY XI, 84

the precipitate thus obtained consists of large crystals, and is therefore more
easily filtered (compare Section XI, 4). It is also of great importance to carry out
the precipitation at boiling temperature, for the relative supersaturation is less at
higher temperatures (compare Section XI, 4). The concentration of hydrochloric
acid is, of course, limited by the solubility of the barium sulphate, but it has been
found that a concentration of 0.05 Af is suitable; the solubility of the precipitate
in the presence of barium chloride at this acidity is negligible. The precipitate may
be washed with cold water, and losses, owing to solubility influences, may be
neglected except for the most accurate work.

Barium sulphate exhibits a remarkable tendency to carry down other salts (see
co-precipitation, Section XI, 5). Whether the results will be low or high will
depend upon the nature of the co-precipitated salt. Thus barium chloride and
barium nitrate are readily co-precipitated. These salts will be an addition to the
true weight of the barium sulphate, hence the results will be high, since the
chloride is unchanged upon ignition and the nitrate will yield barium oxide. The
error due to the chloride will be considerably reduced by the very slow addition of
hot dilute barium chloride solution to the hot sulphate solution, which is
constantly stirred ; that due to the nitrate cannot be avoided, and hence nitrate
ion must always be removed by evaporation with a large excess of hydrochloric
acid before precipitation. Chlorate has a similar effect to nitrate, and is similarly
removed.

Positively charged barium sulphate, which is obtained when sulphate is
precipitated by excess of barium ions, can be coagulated by the addition of a trace
of agar-agar. About 1 mg of agar-agar as a 1 per cent aqueous solution will cause
the flocculation of about 0.1 g of barium sulphate, but in practice somewhat
larger quantities are generally used. The resulting precipitate does not creep up
the sides of the vessel.

Negatively charged barium sulphate, obtained in the determination of barium
IS not appreciably improved by agar-agar; this precipitate as a rule, presents little
difficulty in filtration.

Pure barium sulphate is not decomposed when heated in dry air until a
temperature of about 1400 °C is reached :

BaSO^ =BaO + S 03

T'u

however, easily reduced to sulphide at temperatures above
t- by the carbon of the filter paper :

BaS 04 + 4 C = BaS4-4CO

The reduction is avoided by first charring the paper without inflaming, and then

505

XI, 84 QUANTITATIVE INORGANIC ANALYSIS

burning ofl'thc carbon slowly at a low temperature witli free access of air. If a
reduceti precipitate is obtained, it may be rc-o.xidi-sed by Ireatmenl with sulphuric
acid, followed by volatilisation of the acid and re-heating. The final ignition of
the barium sulphate need not be made at a higher temperature than 600-800 °C
(dull red heat). A Vitreosil or porcelain filtering crucible may be msed, and the
difliculty of reduction by carbon is entirely avoided.

Procetiuri'. Weigh out accurately about 0.3 g of the solid* (or a sufficient
amount to contain 0.05-0.06 g of sulphur) into a 400-cm^ beaker, provided with
a .stirring rod and clock-glass cover. Dissolve the solidt in about 25 cm^ of water,
add 0.3-0.6cm^ of concentrated h>drochIoric acid, and dilute to 200 -225cm\
Heat the solution to boiling, add dropwise from a burette or pipette 10-12cm^ of
wann 5 percent barium chloride solution (5 g BaCl2.2H,0 in 100 cm-^ of water-
ed/. 0.23/). Stir the solution constantly during the addition. Allow the precipitate
to settle for a minute or two. Then test the supernatant liquid for complete
precipitation by adding a few drops of barium chloride solution. If a precipitate is
formed, add slowly a further 3 cm-* of the reagent, allow the precipitate to settle as
before, and test again; repeat this operation until an e.xccss of barium chloride is
present. When an e.sccss of the precipitating agent has been added, keep the
covered solution hot, but not boiling, for an hour (steam balli, low-iemperaiure
hot plate, or .small Dame) in order to allow time for complete precipitation. J The
volume of the .solution should not be allowed to fall below 15t)cm^; if the clock
glass covering the beaker is removed, the under side must be rinsed olTinto the
Ix'aker by means of a .stream of water from a wash bottle. The precipitate should
settle readily, and a clear supernatant liquid should be obtained. Test the latter
with a few' drops of barium chloride solution for complete preeipitation. If no
precipitate is obtained, the barium sulphate is ready for filtration. The
determination may be completed by either of the following processes.

(i) Filter paper method. Decant the clear solution through an ashlcss lilier
paper (Whatman, No. 40 or 540). and collect the filtrate in a clean beaker. Test
the liltraic with a few drops of barium chloride; if a precipitatc'forms, the entire
sample must be di.scarded and a new determination commenced. If no pmeipitate
forms discard the liquid, rinse out the beaker, and place it under the funnel; this is
in order to avoid the necessity vif re-(illering the whole solution if any precipitate
should pass through the filter. Transfer the precipitate to the filter with the aid of
a jet of hot water from the wash bottle. Use a rubber-tipped rod (‘policeman’) to
remove any precipitate adhering to the walls of the beaker or to the stirring rod,
and transfer tlie precipitate to the filler paper. Wash the prccipiuilc with small
portioii-s of hot water. Direct the jet us ne;ir the top of the filter paper as possible,
and let each portion of the wash .solution run through before adding the next.
Continue the washing until about 5 cm-* of the wash solution gives no

* A.R. 1’oUi.imiii iulpliale m.i)' be eiiiplovcJ

t For sulpliaics uhicl) .ire insoluble in ivalcr and.tcids. it is best to mi.v the (inely powdered solid iviih
sis to twelve limes its bulk of anhjdtou'v sodium c.irbonaic in .i plalinum eiucible (Section tit, -15),
heat the covered crucible slowly to fusion, and maintain in the fused stale for 15 minutes. Tlic melt is
extracted with water, the solution filtered, the re.siduc washed with hot I percent sodium carbonate
solution, and the cold filtrate carcfullv acidified wuh hydrochloric ackl (to methyl orange). TIic
sulphate is determined as above.

) An equivalent result is obtained by allowing the solution to stand at the laboratory temperature for

about 18 hours.

506

GRAVIMETRY XI, 85

opalescence with a drop or two of silver nitrate solution. Eight or ten washings
are usually necessary.

Fold the moist paper around the precipitate and place it in a porcelain or silica
(Vitreosil) crucible, previously ignited to redness, cooled in a desiccator and
weighed. Dry the paper by placing the loosely covered crucible upon a triangle
several centimetres above a small flame. Then gradually increase the heat until
the paper chars and volatile matter is expelled. Do not allow the paper to burst
into flame, as mechanical loss may thus ensue. When the charring is complete,
raise the temperature of the crucible to dull redness and burn off the carbon with
free access of air* (crucible slightly inclined with cover displaced, Fig. Ill, 29).
When the precipitate is white,! ignite the crucible at a red heat for 1 0-1 5 minutes.
Then allow the crucible to cool somewhat in the air, transfer it to a desiccator,
and, when cold, weigh the crucible and contents. Repeat the ignition with 10-
minute periods of heating, subsequent cooling in a desiccator, etc., until constant
weight ( ± 0.0002 g) is attained.

Calculate the percentage of SO 4 in the sample.

(ii) Filtering crucible method. Clean, ignite, and weigh either a porcelain
filtering crucible or a Vitreosil filtering crucible (porosity. No. 4). Carry out the
ignition either upon a crucible ignition-dish or by placing the crucible inside a
nickel crucible at a red heat (or, if available, in an electric muffle furnace at
600-800 °C), allow to cool in a desiccator and weigh. Filter the supernatant
liquid, after digestion of' the precipitate, through the weighed crucible, using
gentle suction. Reject the filtrate, after testing for complete precipitation with a
little barium chloride solution. Transfer the precipitate to the crucible and wash
with warm water until 3-5 cm^ of the filtrate give no precipitate with a few drops
of silver nitrate solution. Dry the crucible and precipitate in the oven or at
100-1 10 T, and then ignite in a manner similar to that used for the empty
crucible for periods of 15 minutes until constant weight is attained (1).

Note. 1 . A rapid method for weighing the precipitate is as follows. (This
procedure should not be employed by elementary students or beginners in the
study of quantitative analysis.) Filter off the precipitated barium sulphate
through a weighed filtering crucible (Gooch, sintered glass, or porcelain) and
wash it with hot water until the chloride reaction of the washings is negative.
Then wash five or six times with small volumes of ethanol, followed by five or six
small volumes of anhydrous diethyl ether. Suck the precipitate dry on the pump
for 10 minutes, wipe the outside of the crucible dry with a clean linen cloth, leave
ra a vacuum desiccator for 10 minutes (or until constant in weight), and weigh as
BaSO^. The result is of a moderate order of accuracy.

,85. SULPHIDE. Determination of sulphur in mineral sulphides.

ntroduction. The methods to be described apply to most insoluble sulphides,
snt w ^^Iphtir is oxidised to sulphuric acid, and determined as barium
P a e. Two procedures are available for effecting the oxidation.

Md crucible cover may be removed by placing it, clean side down, on a triangle,

t Ifthr

etc. slightly discoloured, add a drop or two of dilute sulphuric acid, evaporate gently.

507

XI, 84 QUANTITATIVE INORGANIC ANALYSIS

burning off the carbon slowly at u low temperature witli free access of air. [fa
reduced precipitate i.s obtained, it may bo ro-oxidised by treatraeni with sulphuric
acid, followed by volatilisation of liie acid and re-heating. The final ignition of
the barium sulphate need not be made at a higher temperature than 600-S00'C
(dull red heat). A Vitreosii or porcelain filtering crucible may be used, and the
diHiculiy of reduction by carbon is entirely avoided.

Procalwi'. Weigh out accurately about ()..I g of the solid* (or a suflicicni
amount to contain 0.05-0.06 g of sulphur) into a 400-cin^ beaker, pro\ided with
a stirring rod and clock -glas.s cover. Dissolve the solid f in uboiit 25cm^ of water,
add 0..1-0.6cm^ of conceiuraled ludrochloric acid, and dilute to 200“225cra\
i Icat thesolution to boiling, add dropwise from a burette or pipette 10-12cm^ of
warm 5 per cent barium chloride solution (5 g DaCl., 211, Oil! 1 00 cm^ of water—
ca. 0.2.V/). Stir the .solution constantly during the addition. Allow the precipitate
to settle for a minute or two. Then test the supernatant liquid for complete
precipitation by adding a few drops ofbarium chloride solution. If a precipitate is
formed, add slowly a furilier .1 enf’ of the reagent, allow the precipitate to. settle as
before, and test again ; repeat this operation until an excess ofbarium chloride is
present. When an excess of the precipitating agent has been added, keep the
covered solution hot, but not boiling, for an Ijour (steam bath, low-temperature
hot plate, or small llame) in order to allow time for complete precipitation.; The
volume of the solution should not be allowed to fall below ISOcni^; if the clock
glass covering the beaker is removed, the underside must be rinsed off into the
beaker by means of a stream of water from a wash bottle. The precipitate should
settle readily, and a clear supernatant liquid should be obtained. Test the latter
with a few drops of barium chloride solution for complete precipitation, if no
precipitate is obtained, the barium sulphate is ready for filtration. The
determination may be completed by either of tiie following processes.

(i) Filter paper nicihod. Decant the clear solution through an ashless filter
paper (Whatman, No. 40 or 540). and collect the filtrate in a clean beaker. Test
the filtrate svith a few drops of barium chloride; if a precipitaic'forms, the entire
sample must be discarded and a new determination commenced. If no praripiiatc
forms discard the liquid, rin.se out the beaker, and place it under the funnel : this is
in order to avoid the necessity of rc-filicring the whole solution if any precipitate
should puss through the filter. Transfer the precipitate to the filler with the aid of
a jet of hot water from the wash hoide. Use a rubber-tipped rod (‘policeman') to
remove any precipitate adhering to the walls of the beaker or to tbestirring rod,
and transfer the precipitate to the tiller paper. Wash llic precipitate with small
portions of hot water. Direct llic jet as nc.ir the top of the filter paper as possible,
and let each portion of the wash solution run through before adding the ne.xt.
Continue the wa.sliing until about 5ciii^ of the wash solution gives no

* A.H, t’ot.issiuii! sulplulc m;iy be cniplincU

t For sulplulcs wliieh are insoluble iti u jtcrami ueu.lv, ii u hcvl lo mix llie liiiciy pou Jcrcil st'IM "ilb
si.e to twelve limes its bulk of.inlndrous sodium earbon.iio in .i pluiinum crucible iScetion Ill,4>).
1 k.u the covered crucible slowly lo fusion, .ind nuiiiuiii in ilie fujcd suiefor I 5 niinule,->.Themcllis
exiracicd wiih waicr, ihc solution filleted, the- residue washed with hot 1 percent sodium carbiinale
solution, and the cold fillr.ilc c.irefiil!) acidified with hydrochloric acid (tvs methyl orange). The
sulph.ilc is deicrniincd as above

J An equivalent result is obtained by allowing the soliiliun to stand al the hiboralory Icniperatutc for
about 1 8 hours.

S06

GRAVIMETRY XI, 85

opalescence with a drop or two of silver nitrate solution. Eight or ten washings
are usually necessary; _ ■ '

Fold the moist paper around the precipitate and place it in a porcelain or silica
(Vitreosil) crucible, previously, ignited to redness, cooled in a desiccator and
weighed. Dry the paper by placing the loosely covered crucible upon a triangle
several centimetres above a small flame. Then gradually increase the heat until
the paper chars and volatile matter is expelled. Do not allow the paper to burst
into flame, as mechanical loss may thus ensue. When the charring is complete,
raise the temperature of the crucible to dull redness and burn olf the carbon with
free access of air* (crucible slightly inclined with cover displaced. Fig. Ill, 29).
When the precipitate is white, f ignite the crucible at a red heat for 1 0-1 5 minutes.
Then allow the crucible to cool somewhat in the air, transfer it to a desiccator,
and, when cold, weigh the crucible and contents. Repeat the ignition with 10-
minute periods of heating, subsequent cooling in a desiccator, etc., until constant
weight ( + 0.0002 g) is attained.

Calculate the percentage of SO 4 in the sample.

(ii) Filtering crucible method. Clean, ignite, and weigh either a porcelain
filtering crucible or a Vitreosil filtering crucible (porosity. No. 4). Carry out the
ignition either upon a crucible ignition-dish or by placing the crucible inside a
nickel crucible at a red heat (or, if available, in an electric muffle furnace at
600-800 °C), allow to cool in a desiccator and weigh. Filter the supernatant
liquid, after digestion of the precipitate, through the weighed crucible, using
gentle suction. Reject the filtrate, after testing for complete precipitation with a
little barium chloride solution. Transfer the precipitate to the crucible and wash
with warm water until 3-5 cm^ of the filtrate give no precipitate with a few drops
of silver nitrate solution. Dry the crucible and precipitate in the oven or at
100-1 10 °C, and then ignite in a manner similar to that used for the empty
crucible for periods of 15 minutes until constant weight is attained (1).

Note. 1 . A rapid method for weighing the precipitate is as follows. (This
procedure should not be employed by elementary students or beginners in the
study of quantitative analysis.) Filter off the precipitated barium sulphate
through a weighed filtering crucible (Gooch, sintered glass, or porcelain) and
wash it with hot water until the chloride reaction of the washings is negative.
Then wash five or six times with small volumes of ethanol, followed by five or six
small volumes of anhydrous diethyl ether. Suck the precipitate dry on the pump
for 10 minutes, wipe the outside of the crucible dry with a clean linen cloth, leave
R desiccator for 10 minutes (or until constant in weight), and weigh as

BaS 04 . The result is of a moderate order of accuracy.

^.85. SULPHIDE. Determination of sulphur in mineral sulphides.

nti^uction. The methods to be described apply to most insoluble sulphides.
sYh sulphur is oxidised to sulphuric acid, and determined as barium
pnate. Two procedures are available for effecting the oxidation.

and r crucible cover may be removed by placing it, clean side do wn, on a triangle,

for some time.

etc. slightly discoloured, add a drop or two ofdilute sulphuric acid, evaporate gently.

507

XI, 85 QUANTITATIvn INORGANIC ANALYSIS

A. Dry process. Disamion. The oxidation is carried out by fusion with
sodium peroxide, or, less ellicicntly, with sodium carbonate and potassium
nitrate:

2FeS, + I SNa^O, = Fe^j + 4 Na,S 04 + 1 1 Na.O

The sulphide is fused with the sodium peroxide in an iron or nickel crucible
{platinum is strongly attacked— Sections HI, 24 and III, 35). the fused mass
treated with water, filtered, and acidified. The e.xcess peroxide is removed by
boiling, and the sulphate ion precipitated with barium chloride. The
decomposition of the sulphide is rapid, but the method has several
disadvantages. Amongst these may be mentioned ; the slight attack on the metal
crucible, thus preventing the .sub.sequenl determination of tiie metal content of
the sample; the introduction of appreciable quantities of sodium salts, thus
increasing the error due to co-precipitation (Sections XI, 5 and XI, 84); and the
possible contamination by sulphur from the ilame gases, since sulphur dioxide is
rapidly absorbed by the alkaline melt. The last error may be minimised by fitting
the crucible into a hole in a sheet ofa.sbcstos or ‘uraliie’, and keeping the crucible
covered during the ignition (see Section 111,45).

ProMhtre. Dry some finely povvdered pyrites* at 1 00 ‘'C for I hour. Fit an
iron or nickel crucible into a hole in an asbestos or ‘uraliie’ board sullkicnlly
large to allow two-thirds of the crucible to project below the board, Place about
1 g of A.R. anhydrous sodium carbonate into the crucible, and weigh accurately
intoil 0.4 -0,5 g of the pyrites. Add 5- fig of sodium pero.xide, and mis well with a
stout copper or nickel wire or with a thin glass rod. Wipe the wire or rod, if
ncce.ssary, with a small piece of quantitative filter paper, and add the latter to the
crucible; cover the mixture with a thin layer of pero.side. Phice the crucible in the
hole in the asbestos or ‘uraliie’ sheet. ,md heat it with a very small fiamc.
Increase the temperature gradually until after 10 15 minutes the crucible is at a
(lull red heat (the lower the tempcruiurc, the less is tiie crucible attacked) and just
sufficiently hot to keep the mass completely fused. Remove the cover
occasionally and examine the cuiucms; be sure that the whole ma.ss is fluid.
Maintain the mass fluid for 1 5 mimito to complete the oxidation. Allow to cool,
extract the crucible with water in a covered fiOOenv’ beaker, rinse olV the crucible-
cover into the beaker, remove ihecrucibic with a glass rod and wash it welhdilute
to .^OOcm^. Boil the solution for 15 mimucs in order to destroy the exec-ss of
pero.vide (NajOj-l-lll.O = 2NaOII + 11,0.), neutralise part of the alkali by
adding 5-6cm^ of concentrated hydrochloric acid with stirring, add a Whatman
‘accelerator’ or a quarter of an ‘ashlcss tablet anti liller through a W'halinan No.
541 filter paper. Wash the residue at least ten times with hot 1 percent sodium
carbonate solution (10 20-cm-' portions). Acidify the combined filtrate and
washings contained in a 800 lOOOcm^ beaker with concentrated hydrochloric
acid, using methyl red or methyl orange as indicator, anti add 2cin^ of acid in
c.xccss. Dilute, il necessary, to bOOern^, and heal to boiling. Precipitate the
sulphate by the slow addition with stirring of a boiling 5 per cent solution of
barium chloride; the latter is added m slight excess of the calculated amount

liuini Pyriie,, No. -15,10' (one of llic Aii>it),cU g.itnpict for SiuJciU',) atoilablcfroiii die Bure.iuol
Analysed Samplo. is Miiiablc

508

GRAVIMETRY XI, 85

required, assuming the pyrites to be pure FeSj. Complete the determination as in
Section XI, 84.

Calculate the percentage of sulphur in the sample.

B. Wet process. Discussion. The sulphide is oxidised (i) by bromine in
carbon tetrachloride solution, followed by nitric acid, (ii) by sodium chlorate and
hydrochloric acid, or (iii) by a mixture of nitric and hydrochloric acids and a little
bromine. The use of the first-named oxidising agent will be described; the
reaction may be represented by :

2FeS2 + 6HNO3 + ISBrj + IbH^O = 2Fe(N03)3 -I-4H2SO4 -|-30HBr

The method has the advantage of not introducing any metallic ions, but it is
essential to remove the excess of nitric acid (see Discussion in Section XI, 84). The
action is slower than by the fusion method.

Procedure. Dry some finely powdered iron pyrites ( 1) at 100 °C for I hour.
Weigh out accurately 0.4-0.5g of the pyrites into a dry 400-cm^ beaker, add
6cm^ of a mixture of 2 volumes of pure liquid bromine and 3 volumes of pure
carbon tetrachloride (fume cupboard!), and cover with a clock glass. Allow to
stand in the fume cupboard for 15-20 minutes and swirl the contents of the
beaker occsionally during this period. Then add 10 cm^ of concentrated nitric
acid down the side of the beaker, and allow to stand for another 1 5-20 minutes,
swirling occasionally as before. Heat the covered beaker below 100 °C by placing
it on a hot plate or thermostatically controlled water bath until all action has
ceased and most of the bromine has been expelled (about 1 hour). Raise the clock
glass cover by glass hooks resting on the rim of the beaker, or displace it to one
side, and evaporate the liquid to dryness on the steam bath. Add lOcm^
concentrated hydrochloric acid, mix well, and again evaporate to dryness to
eliminate most of the nitric acid. Place the beaker in an oven or in an air bath at

95-100 °C for 30-60 minutes in order to dehydrate any silica which may be
present (2). If the dry residue is heated at a temperature above 100 °C, loss of
sulphuric acid may occur and the determination will be rendered useless. Moisten
the cold, dry residue with 1— 2cm^ of concentrated hydrochloric acid and, after
an interval of 3-5 minutes, dilute with 50 cm^ of hot water, and rinse the sides of
the beaker and the cover glass with water. Digest the contents of the beaker at
100°C for 10 minutes in order to dissolve all soluble salts. Allow the solution to
cool for 5 minutes, and add 0.2-0.3g of aluminium powder to reduce the
iron(III). Gently swirl (or stir) until the solution becomes colourless. Allow to
cool, add a Whatman ‘accelerator’, stir, and rinse down the cover glass and the
sides of the beaker. Filter through a Whatman No. 540 paper, and collect the
mtrate in an 800-cm^ beaker; wash the filter thoroughly with hot water. Dilute
the combined filtrate and washings to 600 cm^ and add 2cm^ of concentrated
ydrochloric acid. Precipitate the sulphate in the cold (3) by running in from a
urette, without stirring, a 5 per cent solution of barium chloride at a rate not
exceeding 5 cm^ per minute until an excess of 5-10 cm^ is present (4). When all
e precipitant has been added, stir gently and allow the precipitate to settle for 2
ours, but preferably overnight. Filter through a No. 540 filter paper or,
pre erably , through a porcelain filtering crucible, wash with warm water until
e rom chloride, and ignite to constant weight as described under A.
calculate the percentage of sulphur in the sample.

0 es. 1. The procedure is applicable to most mineral sulphides; many of

509

XI, 861^7/88 QUAN TITATIVE INORGANIC ANALYSIS

these contain silica, and provision is made for the removal of this impurity in the
experimental details.

2. If the iron pyrites or the .sample of sulphide contains no appreciable
proportion ofsilica. the heating at 95-100 C may be omitted,

3. If a drop or two oftinill) chloride solution i.s added to prevent reo.xidation
of the Fe(IU .salt by air, precipitation of the barium sulphate may be made in
boiling .solution according to the usual procedure (Section XI, 84).

4. Calculate the volume of 5 per cent barium chloride solution which must be
added from the appro.ximale sulphur content of the iron pyrites FeSj or of the
mineral .sulphide.

XI, 86. SULPHITE. Determination of sulphite by oxidation to sulphate and
precipitation as barium sulphate. Disais.uoit. Sulphites may be readily
converted into sulphates by boiling with e.xcc.ss of bromine water, sodium
hypochlorite, sodium hypobroinite. or ammoniacal hydrogen peroxide (equal
volumc.s of 20-volume hydrogen pero.xide and 1 I ammonia solution). The
excess of the reagent is decoinpo.sed by boiling, the solution acidified with
hydrochloric acid, precipitated with barium chloride .solution, and tlie barium
.sulphate collected and weighed in the usual manner (Section .XI, 84).

XI, 87. TIIIOCYaN.VTH This anion may be determined in one of the
following forms.

A. Copper(I) thiocyanate, CuSC.N. The solution (JOOcm^l should be
neutral or slightly acid (hydrochloric or sulphuric acid), and contain not more
than 0.1 g SCN. It is saturated with sulphur dioxide in the cold (or 50cm^ of
freshly prepared .saturated sulphurous acid solution added), and then treated
dropwisc and with constant stirring with about COcur'' of 0.1, if-copper sulphate
solution. The mixture is again saturated with sulphur dioxide (or lOcm^ of
saturated sulphurous acid solution .iddcd), allowed to stand for a few hours,
collected on a weighed filtering crucible (Gooch, sintered gla.ss, or porcelain),
washed .several tirne.s with cold water containing suiphurou.s acid until the copper
IS removed (potassium hexacyanoferrate(U) test), and finally once with ethanol.
The precipitate is dried at 110-120’'C to constant weight, and weiehed as
CuSCN.

B. Barium sulphate, BaS 04 . The ihioeyanale is o.xidiscd with bromine
water to sulphate, and the latter is determined by precipitation as barium
sulphate. All other compounds containing sulphur must be absent. The alkali
thiocyanate solution is treated wiJi excess of bromine water, licatcd for 1 hour
on the water bath, the solution acidified with hydrochloric acid, and the
sulphuric acid precipitated and weighed as BaSO^ (sec Section -XI, 84).

SCN ' +4Br. -1-411,0 == SO^^' -r7Br -b8ir -i-BrCN

XI, 88. THIOSULPlI.VfE. Two methods are commonly used for the
determination of thiosulphates.

A. Conversion of tliio.siilpliate (o sulphate and determination as barium
sulphate. Discussiim. 'Thiosulphates arc oxidised to sulphates by methods
similar to those described for sulphites (Section XI, 86), c.g., by heating on a
water bath with an ammoniacal solution of hydrogen peroxide, followed by
boiling to expel the exce.ss of the reagent. The sulphate is then determined as

510

GRAVIMETRY XI, 89/90

barium sulphate, BaS 04 . One molecule of thiosulphate corresponds to two
molecules of barium sulphate.

B. Determination of thiosulphate as silver sulphide.

Discussion. Thiosulphate may also be determined in almost neutral solution
by the addition of a slight excess of O.lM-silver nitrate solution in the cold
and, after 2-3 minutes, heating at 60 °C in a covered vessel. After cooling,
the precipitate of silver sulphide AgjS is collected, washed with ammonium
nitrate solution, water, and finally with ethanol. The precipitate is dried at 1 10 °C
to constant weight, and weighed as Ag 2 S.

XI, 89. References

1. T. B. Smith (1940). Analytical Processes. 2nd edn. London; Arnold.

2. H. A. Laitinen and W. E. Harris (1975). Chemical Analysis. 2nd edn. New York;
McGraw-Hill.

3. R. Grzeskowiak and T. A. Turner (1973). Talanta, 20, 351.

4. J. Bassett, G. B. Leton and A. I. Vogel (1967). ‘Dioximes of Large Ring 1,2-Diketones
and their Applications to the Determination of Bismuth, Nickel and Palladium’,
Analyst, 92, 279.

XI, 90. Selected bibliography

1. C. J. Rodden (1950). Analytical Chemistry of the Manhatten Project. New York;
McGraw-Hill.

2. R. Fresenius and G. Jander (1940-58). Handbuch der Analytischen Chemie. Dritte
Teil. Bestimmungs und Trennungsmethoden. Berlin; Springer- Verlag.

3. C. Duval (1954-57). Trade de Micro-Analyse Minerale. Vol. I-IV. Paris; Presses
Scientifiques.

4. G. Chariot and D. Bezier (1957). Quantitative Inorganic Analysis (trans. R. C.
Murray). London ; Methuen and Co.

5. L. Gordon, M. L. Salutsky, and H. H. Willard. (1959). Precipitation from Homo-
geneous Solution. New York; John Wiley.

6. C. N. Reilley (ed) (1960). Advances in Analytical Chemistry and Instrumentation. Vol. I.
H. Flashka and A. J. Barnard, Jr. Tetraphenylboron (TPB) as an Analytical
Reagent. New York; Interscience Publishers.

7. F. E. Beamish and J. A. Page. ‘Inorganic Gravimetric and Volumetric Analysis’,
Analytical Chemistry, 1956, 28, 694; 1958, 30, 805.

8. F. E. Beamish and A. D. Westland (1960). ‘Volumetric and Gravimetric Analytical
Methods for Inorganic Compounds’, Analytical Chemistry, 32, 249R.

9. C. L. Wilson and D. W. Wilson (ed.) (1960). Comprehensive Analytical Chemistry.
Vol. lA, Classical Analysis. Amsterdam and London; Elsevier.

10. L. Erdey. Gravimetric Analysis. Part 1 (1963) and Parts 2 and 3 (1965). Oxford;
Pergamon Press.

11. N. H. Furman (ed.) (1962). Standard Methods of Chemical Analysis. Vol. 1, The
Elements, 6th edn. Princeton, New Jersey ; Van Nostrand.

2. Hopkins and Williams (1964). Organic Reagents for Metals and for Certain Radicals.
Vol. II, London.

511

PART E ELECTROANALYTICAL METHODS

513

CHAPTER XII ELECTRO-GRAVIMETRY

XII, 1. theory OF ELECTRO-GRAVIMETRIC ANALYSIS. In electro-
gravimetric analysis the element to be determined is deposited electtolytically
upon a suitable electrode. Filtration is thus avoided, and co-deposition, if the
experimental conditions are carefully controlled, is very rare. The method, when
applicable, has many advantages, and we shall therefore study the theory of the
process in order to understand how and when it may be applied.

Electro-deposition is governed by Ohm’s law and by Faraday’s two laws of
electrolysis (1833-34). The latter state:

1. The amounts of substances liberated at the electrodes of a cell are directly
proportional to the quantity of electricity which passes through the. solution.

2. The amounts of different substances which are deposited, or liberated, by the
same quantity of electricity are proportional to their chemical equivalents:

It follows from the second law that when a given current is passed in series
through solutions containing, say, copper sulphate and silver nitrate
respectively, then the weights of copper and silver deposited will be in the ratio of
theirequivalents, viz., 63.54/2: 107.868.

Ohm’s law expresses the relation between the three fundamental quantities,
current, electromotive force, and resistance.

The current I is directly proportional to the electromotive force E and inversely
proportional to the resistance R, i.e.,

I = E/R

Electrical units. The fundamental SI unit is the unit of current which is
called the ampere and which is defined as the constant current which, if
maintained in two parallel rectilinear conductors of negligible cross section- and
of infinite length and placed one metre- apart in a vacuum, would produce
between these conductors a force equal to 2 x 1 0 “ ^ newton per metre length;

The unit of electrical potential is the volt which is the difference of potential
between two points of a conducting wire which carries a constant current of one
ampere, when the power dissipated between these two points is one watt, or one
joule per second.

The unit of electrical resistance is the ohm, which is the resistance between two
points of a conductor when a constant difference of potential of one volt applied
between these two points produces a current of one ampere. .

Prior to the introduction of the above absolute units in 1948, ‘international’
units were in use ; the relationships between the two sets of units are :

515

XU, J QUANTITATIVE INORG ANtC ANALYSIS

1 international ohm = 1.00049 absolute ohms;

I international ampere - 0.99985 absolute ampere;

1 imernaiional volt - 1.000.VI absolute volts;

for eleclro-giavimetric analysis, the diircrences are insignificant.

The unit quantity ofclectriciiy is the coulomb, ami is tlcfmcd as the quantity of
electricity passing when 1 ampere rtow.s for 1 second. Each coulomb will deposit
l.l 1800 mg of silver.

The weight of an clement liberated by the passage of 1 coulomb of electricity
(or 1 ampere for 1 second) is called the electrochemical equivalent of the element.
The equivalent of silver is I07.S68, hence 107.868/0.00111800, i.e., 96483
coulombs will be required to liberate 1 equivalent of silver. The value generally
employed is 96500 coulombs and thi.s is termed the Faraday coastant(F); this is
the charge associated with one mole of electrons and Inis the accurate value
%4S7Cmor

Some terms used in clcctro-gravlmctric analysis. Voltuic {golvtinic) ami
clcciroiytic ivlls. A ceil consists of two electrodes and one or more solutions in
an appropriate container. If the cell can furnish electrical energy to an external
.sy.stem it is called a volitiic (or yohanic) cell. The chemical energy is converted
more or less completely into electrical energy, but some of the energy may be
dissipated as heat. If the electrical energy is supplied from an c,vtcrna! source the
cell through wltich it Hows is termed an clucirolyiic d'll and Faradav’s laws
account for the m;iierial changes at the electrodes. A given cell may function at
one time as a galvanic cell and at another as an electrolytic cell : a typical c-xample
is the lead accumulatoror sioragcccli. Duringan clcctro-graviraetrieopcrationa
galvanic coll i,s built up as ilie products form on the electrodes. If the current is
switched oUThe products lend to produce a current in a direction opposite to the
direction in which the electrolysis current was passed. The voltage applied to the
electrolysis cell ) must exceed that of the gal vaniccell which is produced (this

is always in oppo.silion to the applied c.m.f. and can be written and must
also overcome the resistance of the solution to the passage of current (i.e., the IR
drop). The amount of current that Hows is given by Ohm’s law:

Cathode. The cathode is the electrode at which reduction occurs. In an
electrolytic cell it is the electrode attached to the negative terminal ofthesoiime,
since electrons leave the source and enter the cleclroIy.sis cell at that terminal. The
cathode is the positive terminal of a galvanic cell, because such a cell aa'epu
electrons at Ibis terminal.

Anode. The tinode is the electrode at which oxidation occurs, It is the
positive terminal of an electrolysis cell or the negative terminal ofa voltaic cell.

Polarised electrode. An electrode is polarised if its potential deviates from
the reversible or equilibrium value. An electrode is said to be depolarised by a
substance if that substance lowens the amount of polarisation.

Current density. The current density is defined as the current per unit area
of electrode surface. It is generally c.xprc.s,sed in amperes per square cm (or per
square dm) of the electrode surface.

Current efficiency. By measuring the amount ofa particular substance that
is deposited and comparing this with the theoretical quantity (calculated by
Faraday’s laws), the actual current cllicicncy may be obtained. In general,

516

r '

|l|l

^ D -

— (v)—

Fig.xn,!

ELECTRO-GRAVIMETRY XII, 1

analytical depositions show low current efficiencies owing to other reactions
which occur during the electrolysis; for example, liberation of hydrogen during
the later stages of the deposition of a metal at a cathode.

Decomposition potential. If a small voltage of, say, 0.5 volt is applied to two
smooth platinum electrodes immersed in a solution of M-sulphuric acid, then an
ammeter placed in the circuit will at first show that an appreciable current is
flowing, but its strength decreases rapidly, and after a short time it becomes
virtually equal to zero. If the applied voltage is gradually increased, there is a
slight increase in the current until, when the applied voltage reaches a certain
value, the current suddenly increases rapidly with increase in the e.m.f. It will be
observed, in general, that at the point at which there is a sudden increase in
current, bubbles of gas commence to be freely evolved at the electrodes. The
experiment may be carried out by means'of the apparatus shown diagrammati-
cally in Fig. XII, 1. A storage battery C is connected across a uniform resistance
wire AB, along which a contact maker D can be moved; the fall of potential
between A and D can thus be varied gradually. Two smooth platinum electrodes
are immersed in M-sulphuric acid in the cell E. V is a suitable voltmeter placed
between the two electrodes of the cell; M is a milliammeter and F is a switch.
When the sliding contact is near to A only a small potential is applied to the
electrodes of the cell; the fall of potential across the cell and the current flowing
through it are read off on the instruments V and M respectively. The applied
voltage is slowly increased by moving D towards B, and the readings of the
voltmeter and ammeter noted after allowing a short time for the values to become
steady. Upon plotting the current against the applied voltage, a curve similar to
that shown in Fig. XII, 2 is obtained; the point at which the current suddenly
increases is evident, and in the instance under consideration is about 1.7 volts.
The voltage at this point is termed the decomposition potential, and it is at this
point that the evolution of both hydrogen and oxygen in the form of bubbles is
first observed. We may define the decomposition potential of an electrolyte as the
minimum external voltage that must be applied in order to bring about
continuous electrolysis.

If the circuit is broken after the e.m.f. has been applied, it will be observed that
the reading on the voltmeter V is at first fairly steady, and then decreases, more or
less rapidly, to zero. The cell E is clearly behaving as a source of current, and is
said to exert a back or counter or polarisation e.m.f., since the latter, acts in a
direction opposite to that of the applied e.m.f. This back e.m.f. arises from the
accumulation of oxygen and hydrogen at the anode and cathode respectively;
two gas electrodes are consequently formed, and the potential difference betweeii
them opposes the applied e.m.f. When the primary current from the battery is
shut off, the cell produces a moderately steady current until the gases at the

517

XU, 2 QUANTITATIVE INORGANIC ANALYSIS

electrodes are either used up or have diffused away; the voltage then falls to zero.
This back e.m.f. is present even when the current from the battery passes through
the cell and accounts for the shape of the curve in Fig. Xlf, 2. It is evident that the
minimum value of the counter e.m.f. may be computed, for it is equal to the
algebraic difference of the electrode potentials which e.xist at the anode and
cathode respectively. This calculation will be referred to again in the succeeding
paragraph.s.

The back e.m.f. is usually regarded as being made up of three components:

(a) The reversible back e.m.f. This i.s the reversible e.m.f of the voltaic cell set
up by the passage of the elect roly tic current, and is based upon concentrations of
solutes in the bulk ofthe solution.

(/i) A concentration polarisation e.m.f or concentration overvoltage. This is
the effect of changes in concentration at an electrode surface with reference to the
concentration of the bulk of the solution. Thus in the electrolysis of an acidic
solution of copper sulphate between platinum electrodes, concentration changes
occur both at the cathode and the anode. At the cathode depletion of copper
ions occurs near the surface; the reversible potential of the copper electrode
therefore shifl.s in the negative direction. At the anode accumuialion of
hydrogen ions (211 jO ->■ O, -r-lU* +4f)and perhaps of o.xygen (ifthesolutionis
not already saturated with it) cuuse.s the reversible potential of the oxygen
cleeirode to shift in the positive direction. Both effects tend to increase the back
e.m.f The concentration overvoltage is increased by increased current density
and decreased by stirring.

(e) An activation overvoltage. This is ilie departure of the potential of an
electrode from its reversible value due to tiie pa.ssage ofthe electrolytic current.

It is observed that whereas the decomposition potentials ofsalt solutions vary
considerably, those for acids and alkalis, with the exception ofthe halogen acids,
are all approximately 1 .7 volts. It is therefore concluded that the same electrolytic
proce.ss occurs with these acids and bases; this can only be the evolution of
hydrogen at the cathode and of oxygen at the anode:

2H^ -j-2t’^e: 11, (acidic mediuni)

21110 + 2es?i H, + 20H ' (basic mediunt)

2H ,0 Oi -t-4H *■ 4- 4<' (acidic medium)

40H ^ O , + 2HjO + 4e (basic medium)

Tile net ceil reaction is the decomposition orwaier;

2H,0 = 2Hi + 0,

With the halogen acids in .1/ .solution, the halogen and not oxygen is liberated at
the anode, since the drscharge ofthe halogen ion can occur more readily than that
of hydroxide ion ; the discharge potential varies with the halogen.

For a similar electrolysis of M-iinc sulphate solution, the reactions at the
cathode and anode are respectively:

Zn-'' + 2e^Zn
2H,0^0,+4H‘ +4e

an o.xygen electrode being produced at the anode.

Xll, 2. ELECTRODE REACTIONS. In electro-gravimetric analysis we are
largely concerned with the electrolysis of salt solutions', and it is therefore

518

ELECTRO-GRAVIMETRY XU, 3

proposed to study in some detail the reactions which take place at the electrodes.
Let us consider the electrolysis of a molar solution* of zinc broimde between
smooth platinum electrodes. The application of a voltage will result
deposition of zinc on the cathode (thus producing a zinc electrode) and of
bromine at the anode (thus producing a bromine electrode). The reaction at the
cathode is : ' ,

Zn2+-l-2e:?iZn

i.e., a reduction (Section 11, 23), and that at the anode is :

2Br”:?iBr2-l-2e

i.e., an oxidation. Thus reduction occurs at the cathode and oxidation at the
anode. We may calculate the potential at the cathode at 25 °C from the formula
(Section 11, 20) :

= £^n+^log [Zn^"] =
since [Zn^*] = 1 mole dm“^.

At the anode;

^ 0.0591, „ -1 E-e 0.0591, .

j— log[Brl=£^Br, p-log2

since [Br“] = 2moledm“^.

The e.m.f. of the resulting cell will therefore be :

£®zn- ^^log2} = 0.76-(- 1.07-0.02) = 1.85 volts

In general, it may be stated that the theoretical back or polarisation e.m.f. Ei^ck
is given by;

^bock ^cathode ^anode

where Eca,hode and are calculated as already described (Section II, 20).

Xn, 3. OVERPOTENTIAL. It has been found by experinient that the
decomposition voltage of an electrolyte varies with the nature of the electrodes
employed for the electrolysis and is, in many instances, higher than that
computed from the difference of the reversible electrode potentials. The excess
voltage over the calculated back e.m.f. is termed the overvoltage.! Overpotential
may occur at the anode as well as at the cathode. The decomposition voltage E^ is
therefore;

^cathode + -E'o.c. ~ (■E'ano* + -£0.0.)

where ^ and Eo,a. are the overpotentials at the cathode and anode respectively.

The overpotential at the anode or cathode is a function of the, following
variables; , ,, ,

1. The nature and the physical state of the metal employed for the electrodes.

* Le. molar with respect to Zn^ + . Ideally, concentration should be replaced by activity ; the former is
however, sufficiently accurate for our purpose. , *

t The term overvoltage should strictly be applied to a cell, and overpotential to a single electrode.

519

XII, 3 QUANTITATIVE INORGANIC ANALYSIS

The fact that reactions involving gas evolution usually require less ovcrvoltageat
platinised tiian at polished platinum electrodes is due to the much larger area of
the platinised electrode and the smaller current density at a given electrolysis
current.

2. The physical state of the substance deposited. If it is a metal, the
overpoienlial is usually small; if it is a gas, such as o.sygen or hydrogen, the
overpotential is relatively great.

3. The current density employed. For current densities up to 0.0! ampere
cm ■' % the increase in overpotential is very rapid ; above thi.s figure the increase in
overpotential continues, but less rapidly.

4. The change in concentration, or the concentration gradient, existing in the
immediate vicinity of the electrodes ; as this incrca.ses, the overpotential ri.ses. The
concentration gradient depends upon the current density, the temperature, and
the rate of stirring of the solution.

5. The overpotential decreases, often very considerably, with increasing
temperature.

The oserpoteniial of hydrogen is of great importance in electrolytic
determinations and sep.irations. Some values are collected in Table XII, I.

Table XII, I Hydrogen overpotential on various cathodes (in volts)

Calhoilc

Solution

HrM visible Cufrcnl dcusily
gas bubblo

0 01 Jiiip etii ■ *

0. 1 jnip. cHi"'

i.tMl.SO,

0.00

0.09

0.16

l.U-H.SO.,

0 02

0.4

1.0

1,U41..S0,

0,10

o..t

0 9

0 07

0.2

l.t/-ll.SO*

0.11

0,3

0,7

0 10

0.4

III

i.lMLSO*

0 4

0 05.U-H.SO.

O.dO

0 5

1.2

O.UOS.W-lijSO,

0 4

1.2

t.U-H;SO,

0 04

0.4

1.2

l.VMtjSO,

0.5

1.1

t.U-H.SO*

1.2

1.3

The hydrogen overpotential is greatest with the relatively soft mctal.s, such as
bismuth, cadmium, tin, lead, and zinc, and especially mercury (in the hast case the
value is about 1 .0 volt, but is dependent on the current density). The e.vistence ot
hydrogen overpotential renders possible the electro-gravimetric determination
of metals, such as cadmium and zinc, which otherwise could not be deposited
before tlie reduction of liydrogen ion. In alkaline solution, the hydrogen
overpotential is slightly higher (0.05-0.3 volt) than in acid solution.

The o.xygen overpotential is about 0.4-().5 volt at a polished platinum anode in
acid solution, and is of the order of I volt in alkaline solution with current
densities of 0.02-0.03 amp. cm ' *. As a rule the overpotential a.ssociatcd with the
deposition of some metals ( Ag, Hg. Cu. Pb. Cd. and Zn) on the cathode is quite
small (about 0. 1-0.3 volt) becau.se the depositions proceed nearly reversibly.
When the depositions do not proceed reversibly the overpotential may attain the
same order ot magnitude as for hydrogen evolution. Thus nickel ions show an
overpotential of about 0.6 volt at a mercury cathode.

520

ELECTRO-GRAVIMETRY XII, 4

XII, 4. COMPLETENESS OF DEPOSITION. ■ The voltage E,pp,, applied to
an electrolytic cell must overcome the decomposition potential Eo or back e.m.f.,
as well as the ohmic resistance of the solution, i.e., must be equal, to or
greater than E^+IR- It has been shown (Section XII, 3) that :

Ep — Efyf.' “b .... ' . ,

where E^e and ^anode are the reversible cathode and anode potentials
respectively, is the overpotential effect at the cathode and* is the
overpotential effect at the anode. Overpotential at the cathode makes the
effective cathode potential more negative, than the equilibrium value, and at
the anode causes the effective anode potential to be more positive. Let us consider
the variations in e.m.f. at the cathode during the deposition of a metal in an
electrolytic determination. Let the ionic concentration at the commencement of
the estimation be q. For a bivalent metal, e.g., copper, the cathode potential at
25 °C will be: ^ ‘ '

F^^^logc,. = E%n +0.0296 log q. volts

If the ionic concentration is reduced by deposition to one-ten-thousandth of its
original value (i.e., to secure an accuracy of 0.01 per cent in the determination),
the new cathode potential will be :

+0.0296 log(qxl0-^)

= E%u +0.0296 log c, + 0.0296 log lO"'^

= (£® M" + 0.0296 log q) - 4 x 0.0296
= Potential at commencement of deposition — 0.1 1 8 volt

This reduction in potential is independent of the value of q-, and hence
whenever the ionic concentration is reduced to one-ten-thousandth of its initial
value (this may be regarded as the ultimate limit of an electro-gravimetric
determination, although for most purposes an accuracy of 0.1 per cent is
regarded as sufficient), the potential is altered by 4 x 0.0591/2 = 0.1 18 volt for a
bivalent ion. For a univalent ion, the change is 4 x 0.059 1 / 1 =0.236 volt, and for
a trivalent ion it is 4 x 0.0591/3 = 0.079 volt. Since the back e.m.f. is produced by
a metal cathode acting as the negative pole of a cell, the positive pole being, say,
oxygen, it follows that the back e.m.f. will become greater as the cathode;
becomes more negative during the course of the analysis. Otherwise expressed,
the decomposition potential increases as the deposition of the metal proceeds.
For quantitative deposition the applied e.m.f. must equal or exceed the
decomposition voltage when the concentration of the given cation is negligibly
small (say 1 0 “ of the initial value).

It is important to know the conditions for the deposition of the metal in
preference to hydrogen in an electrolysis. The condition for the deposition of the
metal is evidently that the potential difference between the electrolyte and the
cathode, must be less than the reversible deposition potential of

hydrogen plus the overpotenial of hydrogen (©„,) for the metal under

consideration. The relationship may be expressed in several ways :

RioluUon, metal <(£h + %,),

< (1 — +Oh,),

< (0.059 pH + Oh^) (at 25 °C)

521

XII, 5 QUANTITATIVE INORGANIC ANALYSIS

Xil.5. ELECrROLYTIC SEPARATION OFiVIETALS. When a constant
current is passed through a solution containing two or more electrolytes the
electrochemical process with the most positive reduction potential will occur first
at the cathode, followed by the nc.\t most positive electrochemical process, etc.
Thus if a current is passed through a solution containing copper, hydrogen and
cadmium ions, copper will be deposited first at the cathode. As the copper
deposits, the electrode potential decreases, and when the potential is equal to that
of the hydrogen ions hydrogen gas will form at the cathode. Tlie potential at the
cathode will remain virtually constant as long us hydrogen is evolved, which is
usually until all the water is electrolysed; the potential of the cathode cannot
therefore become sutliciently negative to cause the deposition of cadmium ions
(see Section H, 20). Thus metal ions with positive reduction potentials m;iy be
separated, without external control of the cathode potential, from metal ions
having negative reduction potentials. In practice, tlie hydrogen overpotential on
the cathode plus the rev crsible reduction potential of the hydrogen ions must be
less than the negative reduction potential of any metal ions that remain in
solution. For c.x.amplc, copper ions in a solution containing I.U-hydrogen ions
may be separated from all metallic ions whose reduction potentials are more
negative than about -0.8 volt (the Ivydrogen ovcrpoiential on a copper
electrode); the reversible reduction potential of hydrogen ions is 0.0 volt in this
medium. The principle of separating, by elcctro-deposiiion at constant current,
metallic ions whose reduction potentials arc on dilTcrcnt sides of the potential of
the hydrogen electrode fmd.s application for analytical .separations and
determinations. It will be clear from what has been .stated in the preceding
Section that the initial licpusition potentials of two metals must dificr by at least
0.25 volt for a virtually quantitative separation to be theoretically pos,sible. This
minimum value would require a very precise control of the potential drop at the
cathode; for most priiciical purposes, ihedilference in potential should be at least
0.4 volt. (The procedure for controlled cathode poicnliat is discussed in Sections
XII, 7 14.) Certain metals can be separated electrolyticaliy with great ease, for
e.xample, copper from zinc, nickel, ami cobalt, silver from copper, etc; when,
however, the standard potentials of the two metals dilTcr only slightly, the
electro-separation is more dillicult. The obvious method is to alter the electrode
potential of one of the metals in some way. This is most simply achieved by
decreasing the ionic concentration of theion beingdischarged by incorporating it
in a complex ion of large stability con.stam (Section 11, 11). The deposition
potential of the metal forming a complex ion is thus raised. Furthermore, the
ovcrpoiential at the small ionic conccntniiion is also usually increased. A
possible consequence of changing the electrode potential in this w;iy is that a
metal which in simple ionic solution is liberated at a lower voltage than another
metal, may exhibit the reverse behaviour in a complex-forming medium.

Some results for the deposition potentials (~ - of some metals in
simple and alkali cyanide solutions are given in Table xi l, 2.

An interesting application of these results is to the direct quantitative
separation of copper and cadmium. Tlie copper is first deposited in acid solution ;
the solution is then neutralised with pure aqueous sodium hydroxide, potassium
cyanide is added until the initial precipitate just re-dissolves, and the cadmium is
deposited clcctrolytically. Another application is to the separation of copper and
bismuth: these two metals cannot be separated clcctrolytically from solutions of
their simple salts. If cyanide is added, the copper ions form a cyano complex, and

522

ELECTRO-GRAVIMETRY XII, 6

Table Xn, 2 Deposition potentials’££, of some metals in simple
and in alkali cyanide solutions

Metal

BoforO.lM
solutions of.
ions (volt)

Concentration of excess of KCN
per 0.1 mol. of simple metallic
cyanide

0.2M

0.4M

+0.79

+ 1.18

+ 1.23

+ 0.44

+ 0.71

+ 0.87

+ 0.90

-0.31

+ 0.61

+ 0.96

+ 1.17

the deposition potential is much more negative than before; the bismuth-ion
concentration and the electrode potential are hardly affected and a separation
from copper becomes possible, the bismuth depositing first. The separation is
improved if tartrate is also added.

XII, 6. CHARACTER OF THE DEPOSIT. The ideal deposit for analytical
purposes is adherent, dense, and smooth; in this form it is readily washed without
loss. Flaky, spongy, powdery, or granular deposits adhere only loosely to the
electrode, and for this and other reasons should be avoided.

As a rule, more satisfactory deposits are obtained when the metal is deposited
from a solution in which it is present as complex ions rather than as simple ions.
Thus silver is obtained in a more adherent form from a solution containing the
[Ag(CN2] ion than from silver nitrate solution. Nickel when deposited from
solutions containing the complex ion [Ni(NH3)g]^ ''' is in a very satisfactory state
for drying and weighing. Mechanical stirring often improves the character of the
deposit, since large changes of concentration at the electrode are reduced, i.e.,

concentration polarisation is brought to a minimum.

Increased current density up to a certain critical value leads to a diminution of
grain size of the deposit. Beyond this value, which depends inter alia upon the
nature of the electrol^e, the rate of stirring, and the temperature, the deposits
tend to become unsatisfactory. At sufficiently high values of the current density,
evolution of hydrogen may occur owing to the depletion of metal ions near the
cathode. If appreciable evolution of hydrogen occurs, the deposit will usually
become broken up and irregular; spongy and poorly adherent deposits are
generally obtained under such conditions. For this reason the addition of nitric
acid or ammonium nitrate is often recommended in the determination of certain
metals, such as copper; bubble formation is thus considerably reduced. The
action of the nitrate ion at the copper cathode can be represented by ;

NO3- -H 0 H+ -t-8e = NH4 + -fSHjO

The nitrate ion is reduced to ammonium ion at a lower (i.e,, less negative)
cathode potential than that at which hydrogen ion is discharged, and therefore
Spolarisw hydrogen evolution. The nitrate ion acts as- a cathodic

Raising the temperature, say, to between 70 and 80 “C. often improves the

deSase^rrP^^T^ factors, which include the

ecrease m resistance of the solution, increased rate of stirring and of diffusion

Skes DoShlel? Stirring by heat or by Lchanical mean;

makes possible the use of a high current density and therefore a rapid deposition

523

XU, 7 QUANTITATIVE INORGANIC ANALYSIS

In practice, two methods of electrolysis are utilised. In the first method
stationary electrodes are used and the solution is not .stirred; small current
densities must of necessity be applied in order to secure a coherent depo.sit, and
the procedure is a slow one (slow eleetroly.sLs). In the second method, which ha,s
largely superseded the first, the solution is rapidly stirred (rapid electrolysis).
Various devices are employed for stirring. An independent mechanical stirrer
may be used, but it is more usual to have a rotating anode, which may comsist,
after Sand (Ref. 7), of a platinum gau/e cylinder .stirrounded by a similar (but
stationary) cylinder, which constitutes the cathode, tlie intervening space being
small (3-5 mm). A very much higher current density may then be applied without
seriously affecting the purity or the physical character of the deposit. The .stirring
results in a lilreral supply of metal ions always being present near the cathode,
and consequently the current is principally used in the deposition of the metal. A
considerable saving of time is lints etfecicd, and this accounts for the popularity
of the method. It must be empha.sised that when tlie electrolysis is complete, the
current must not be switched oil as long as the electrodes are in the solution. If the
circuit were broken, the counter c.m.f. would come into play, and this would
cause part of the metallic deposit to pass back into solution.

XII, 7. ELECI'ROLYITC SEPARATION OF .METAlii WITH CON-
TROLLED CATHODE POTENTIAL. For electrolysis to proceed the
c.m.f. which must Ik* applied to an aqueous solution of an electrolyte is given by
the exprc-vsion (compare Section XII, 4):

1 w 'i- F,,,. - (F,...,j^ -I- ) + //?

In tlie common method of clcctro-gravimciric analysis, a voltage slightly greater
than E[, is applied and the electrolysis is allowed to proceed without further
attention, except perhaps to occasionally increase the applied voltage to
maintain the current at appro.vimaicly the same v.due. This process, termed
coii-vtant current electrolysis, when .ipplied to the separation of metals is limited to
the separation of those metals below itydrogen in iltc electrochemical scries from
those above hydrogen. Following tlie deposition of the fust metal (tlie one lower
in the series) hydrogen is evolved at the cathode, and as long as the solution
remains acid the second metal is not deposited. This is e.vcmplified by the
separation of copper from nickel, and from zinc in a sulphuric acid solution. If
the second metal lies only .slightly above the other in the electrochemical series,
separation is no longer possible unless the decompo.siliou potentials can be
displaced eitiier through tlic formation ofan appropriate complex ion or by other
means (compare Section XII, 5). The separation of .such a mixture may be
ell'ected by the application of controlled cathode potential electrolysis. .Xn
auxiliary standard electrode (which may be a saturated calomel electrode with
the tip of the salt bridge very close to the cathode or working electrode) is inserted
in the solution, and thu.s the voltage between the cathode and the reference half-
cell may be mca.sured. It is thus po.ssibIe to isolate ihecllcet at the cathode and to
limit the potential tit this electrode during the electrolysis to a definite value by
decreasing the overall voltage applied to the cathode and anode. We have already
seen (Section XII, 4) that every ten-fold decrease in metal-ion concentration
makes the cathode potential 0.059 1 /« volt more negative at 25 ”C (« is the valency
of the ion). For an accuracy of O.I per cent, the concentration of the ion is
reduced to 10“^ of the original value, consequently the potential will decrease by

524

ELECTRO-GRAVIMETRY Xn, 7

3 X 0.0591/« volt, i.e., 0.177 volt for a univalent ion, 0.088 volt for a divalent ion,
etc. Thus, by controlling the cathode potential with the aid of an auxiliary
electrode, the separation of one metal from another lying somewhat higher in the
electrochemical series becomes possible. Manual control of the potential may
become tedious except for occasional determinations, but the time may be
materially reduced by the use of high current densities ; however methods for the

automatic, control of :cathode potential
have been developed (see Refs. 1-4). •

A simple circuit and apparatus for
controlled cathode potential electror
analysis is shown in Fig. XII, 3;'this will
serve to illustrate the principles of the
technique involved. The various com-
ponents of the apparatus are: a source
of current, which may be a large storage
battery ; a saturated calomel electrode ; a
voltmeter P to indicate the e.m.f. applied
to the cell for the electrolysis; an ad-
justable resistance R capable of carry-
ing current up to 10-15 amps.; a
platinum-gauze cathode C; and a
platinum-gauze anode A which can be
rotated. The potential between the
saturated calomel electrode and the
cathode must be measured with an
instrument which draws little or neglig-
ible current from the reference cell: a
digital voltmeter V is satisfactory. The
total potential measured E is equal to the difference between the potentials of the
calomel electrode and the cathode :

^ ^cal. sat. (.^cathode 4 " }

Since is known, the electrode potential of the cathode can be easily referred
to the hydrogen scale. In order to prevent the cathode potential from exceeding a
fixed value, it is simply necessary to decrease the potential applied to the cathode
and anode by increasing the value of the resistance R.

It may be noted that in evaluating the limiting cathode potential to effect the
separation of one metal from another, a simple computation of the equilibrium
potential from the Nernst equation is insufficient ; the equilibrium potential must
be increased by the overpotential. The latter depends, inter alia, upon the rate of
stirring, the current density, the temperature, as well as upon the nature of the
metal surface ; in consequence, the limiting potential must be established by other
methods. ...

One procedure utilises current-working electrode potential curves for the two
substances to be separated. The current-electrode potential curve is determined
for each reaction under exactly the same conditions that will prevail in the actual
analysis. The potential of the working electrode is increased in regular increments
by increasing the total voltage applied to the cell and is measured against a
standard reference electrode (usually a saturated calomel electrode). The current
IS observed at each value of the electrode potential: to minimise change in the

Fig.Xn,3

525

Cufictit

XU, 7 QUANTITATIVE INORGANIC ANALYSIS

cornposiiion of the solution, especially when the current is large, the cell circuit
should be closed only long enough to make the current measurement at each
value of the electrode potential. Schematic current-cathode potential curves for
the reduction of two substances X and Y areshown in Fig. XII, 4. To initiate the
deposition of the substance X. the cathode potential £, must be at least as large as
the value of the ‘decomposition’ potential indicated by x, but the potential
should not exceed y, for if it docs Y will deposit also. Consequently, for the
complete deposition of X, the cathode potential should be limited to a value
slightly less than that which corresponds to the potential y. The initial current
should not exceed the value indicated by r. As the deposition of the substance X
proceeds, the cathode potential tends to become more negative but is prevented
from e.xcccding the value y by decreasing the voltage applied to the cell.

The course of the current and of the total applied voltage during a typical
controlled potential electro-deposition of copper Ls shown in Fig. XII. 5 (due to
Lingane). It concerns the deposition of 0.2 g of copper at a platinum cathode
from an acid tartrate solution (pH - 4.5). in the presence of hydrazine as anodic
dcpolariser, with the catliode potential maintained at - 0.36x0.02 volt rs. S.C.E.

The initial value of the current at 2.6 amp decreased to 0.03 amp after 17
minutes and finally fell to less than O.Ol amp after 30 minutes. The potentiostat
continuously decreased the total applied voltage from 2.2 volts to 0.48 volt to
maintain the cathode potential constant.

In general, it may be stated that the control potential (constant potential of the
working electrode) for a iwo-clcclron reaction need be no more than 0.15-0.20
volt greater than the decomposition potential to obtain rapid electrolysis. For a
reversible one-electron reaction, the difference between the decomposition
voltage and the control potential must be twice as great as for a two-electron
reaction for equally complete deposition: if 99.9 per cent complete reaction is
acceptable a difference of 0.18 volt would sullicc.

Ordinary polarograin:-.r.>:;uncd with the dropping mercury electrode (see
Chapter XVI) may be u;;,. ‘ | uc the optimum control potential for controlled

526

Applied e,m.f.

ELECTRO-GRAVIMETRY XII, 8

potential electro-deposition with a large mercury working electrode. The
potential corresponds to that at the top of the corresponding polarographic wave
where the diffusion plateau begins: usually it is about 0.1-0.15 volt beyond the
polarographic half-wave potential (compare Section XVT, 3).

Anodic re-oxidation of the metal if it can exist in more than one valency state,
or any reaction between the plated metal and the anodic oxidation, products,
must be reduced to a minimum for trustworthy results, and also to minimise the
time required for the deposition. This may be achieved by such methods as (i) the
use of a reducing agent which will be oxidised in preference to the intermediate
valency state, i.e., an anodic depolariser (e.g., a hydraziniiim salt,
N 2 H 5 ^ = -f 4e, or hydroxylamine, largely 2 NH 2 OH =

1420-1-411^-1-1120+46), (ii) isolation of the anode by means of a membrane,
porous cup, or its equivalent, and (iii) reduction of the anode potential
to a value which will not oxidise the ion in the intermediate valency state.

Electrolytic determinations at constant current

XII, 8. APPARATUS. A. Electrolysis unit The actual set-up employed
will vary from one laboratory to another; a. simple circuit, employing the d.c.
mains (200-240 volts or 110 volts), is shown in Fig. XII, 6. M is the d.c. mains,
Rj is a fixed resistance (which may consist of a bank of lamps), R 2 is a small
high wattage, variable resistance, A is an ammeter reading up to 10 amps, V is
a voltmeter reading up to 10-15 volts, E is an electrolysis vessel, and S is a switch.

Alternatively, the source of d.c. may
be a large-capacity 6-volt or 12-volt
car battery, or a number of accuniu-
lators connected in series. If a d.c.
mains supply is not at hand, any of
the commercial d.b. power supply
units operating frorh the a.c. mains
may be used. In these a'transformer
steps the voltage down to 3-15
volts, the current is then passed
through a rectifier, and finally
through a smoothing filter circuit.

If the polarities of the terminals are not known, they may be determined by
touching the two wires from the terminals on to paper moistened with potassium
iodide solution : a brown stain of iodine will form at the positive pole.

Many types of commercial apparatus for electrolytic analysis are available;
Fig. XII, 7 shows the B.T.L. apparatus (Baird and Tatlock Ltd). This is designed
for use on 200-250 volt a.c. mains and incorporates silicon rectifiers providing
low tension direct current for the electrolysis. Stirring is accomphshed by means
of a magnetic stirrer mounted beneath the stainless steel drip tray on which, the
beaker containing the solution to be electrolysed, is placed. The electrode holders
are mounted in an arm which can be adjusted for height.

B. Electrodes. These are generally made of platinum or of platinum-
iridium, although platinum-coated titanium electrodes are available (Baird and
Tatlock) : gauze electrodes are preferred as they assist the circulation of the
solution and thus help to reduce any tendency to local depletion of the
electrolyte. Typical electrodes are shown in Fig. XII, 8: (a) and (3) represent a

527

Current

XU, 7 QUANTITATIVE INORGANIC ANALYSIS

composition of the solution, especially when the current is large, the cell circuit
should be closed only long enough to make the current measurement at each
value of the electrode polcntial. Schematic current-cathode potential curves for
the reduction of two substances X and Y are shown in Fig. XII, 4. To initiate the
deposition of thesubstanceX, the cathode potential E, must be at least as large as
the value of the ‘decomposition' potential indicated by .v, but the potential
should not c.xcced y, for if it does Y will deposit also. Consequently, for the
complete deposition of X, the cathode potential should be limited to a value
slightly less than that wliich corresponds to the potential y. The initial current
should not exceed the value indicated by r. As the deposition of thesub.stanceX
proceeds, the cathode potential tends to become more negative but is prevented
from e.xcccding the value y by decreasing the voltage applied to the cell.

The course of the current and of the total applied voltage during a typical
controlled potential electro-deposition of copper is shown in Fig. XII, 5 {due to
Linganc). It concerns liic depo.sition of 0.2 g of copper at a platinum cathode
from an acid tartrate solution (pH — 4.5), in the presence of hydrazine as anodic
dcpolariser, with the cathode potential maintained at -0.36±0.02 volt rs. S.C.E

The initial value of the current at 2.6 amp decreased to 0.03 amp after 17
minutes and finally fell to less than 0.01 amp after 30 minutes. The potentiostat
continuously decreased the total applied voltage from 2.2 volts to 0.48 volt to
maintain the cathode polcntial constant.

In general, it may be stated that the control potential (constant potential of the
working electrode) for a iwo-elcclron reaction need be no more than 0.15-0.20
volt greater than the decomposition potential to obtain rapid electrolysis. For a
reversible one-electron reaction, the diHercnce between the decomposition
voltage and the control potential must be twice as great as for a two-electron
reaction for equally complete deposition: if 99.9 per cent complete reaction is
acceptable a dilTerence ofO. 1 8 volt would suHicc.

Ordinary polarograms obtained with the dropping mercury electrode (see
Chapter XVI) may be used to define thcoptinium control potential for controlled

526

Applied c.m.r, CvtiUi)

ELECTRO-GRAVIMETRY XII, 8

.platinum surface, the estimation of the surface area presents no particular
difficulty. As a rough approximation the usual gauze electrode may be regarded
as having an effective area twice that of a plain foil electrode of the same
dimensions.- If an accurate value is required, the actual surface area of the
electrode material must be calculated: The total length of wire can be calculated
from the number of meshes and the dimensions of the electrode. The effective
area will then be the total length of the wire multiplied by %d, where d is the
diameter of the wire.

A mercury cathode finds widespread application for separations by constant
current electrolysis. The most important use is the separation of the alkali and
alkaline-earth metals, Al, Be, Mg, Ta, V, Zr, W, U, and the Lanthanoids from
such elements as Fe, Cr, Ni, Co, Zn; Mo, Cd, Cu, Sn, Bi, Ag, Ge, Pd, Pt, Au, Rh,
Ir, and Tl, which can, under suitable conditions, be deposited on a mercury
cathode. The method is therefore of particular value for the determination of Al,
etc, in steels and alloys ; it is also applied in the separation of iron from such
elements as titanium, vanadium, and uranium. In an uncontrolled constant-
current electrolysis in an acid medium the cathode potential is limited by the
potential at which hydrogen ion is reduced; the overvoltage of hydrogen on
mercury is high (about 0.8 volt), and consequently more metals are deposited
from an acid solution at a mercury cathode than with a platinum cathode. Four
types of mercury celt are shown in Fig. XII, 9. In (a) the platinum wire is sealed
into the side of a lipped Pyrex beaker (250 cm^), whilst in {b). the platinum wire is
sealed into the side tube; the latter type permits the almost complete separation
of the aqueous and mercury layers. Apparatus (c) is perhaps the most useful
form; the diagram is almost self-explanatory. The electrolysis vessel contains the
platinum anode (preferably of the rotating type) immersed in the electrolyte.

<d)

Fig.Xn,9

Electrical contact to the mercury is made through a platinum wire sealed into the
side of the vessel (alternatively, a piece of amalgamated copper wire dipping into
mercury contained in a glass tube, into the lower end of which a short platinum
wire IS sealed, may be used for electrical contact); the mercury acts as cathode:
the stirnng should agitate both the mercury and the solution. When electrolysis is

529

ELECTRO-GRAVIMETRY XU, -8

fitted into two jaws at the end of the metal bar
and is fixed in position by the clamping screw B.
The speed of the motor is controlled by the
adjustable resistance on the base, of the stand.
The direct current may be obtained from the
laboratory d.c. supply, or from the terminals of a
commercial electro-analyser designed for stirring
of the electrolyte with a ^ass or magnetic stirrer.

E. Use and care of electrodes. Electrodes
must be free from grease, otherwise an adherent
deposit may not be obtained. For this reason an
electrode, should never be touched on the
deposition surface with the fingers; it should
always be handled by the platinum wire or rod
attached to the main body of the electrode.
Platinum electrodes are easily rendered grease
free by heating them to redness in a flame.

Before use electrodes must be carefully cleaned
to remove any previous deposits. Deposits of
copper, silver, cadmium, mercury, and many
other metals can be removed by immersion in
dilute nitric acid (1:1), rinsing with water,
then boiling with fresh 1:1 nitric acid for

5-10 minutes, followed by a final washing with
water. Copper may also be removed by means of a solution composed of 20 g
trichloroacetic acid, 100 cm^ concentrated ammonia solution, and 100 cm^

water. Deposits of lead dioxide are best removed by means of 1 1 1 nitric acid
containing a little hydrogen peroxide to reduce the lead to the bivalent form ;
ethanol or oxalic acid may replace the hydrogen peroxide. In all cases it is
recommended that the electrodes be heated to bright redness over the colourless
flame of a Bunsen burner before use.

When the electrolysis is complete, stirring is stopped, and the electrolysis
beaker (where possible) is lowered away from the electrodes before breaking the
circuit] the latter is necessary, since otherwise the electrolyte in contact with the
electrode may dissolve some of the deposit. The electrodes are washed
immediately with a fine stream of distill^*d water directed uniformly around the
upper rims of both electrodes from a wash bottle: the first 10-15 cm^ will then
contain virtually all the electrolyte adhering to the electrodes. It is unnecessary to
save subsequent washings, thus avoiding excessive dilution of the residual
solution. The electrode is then disconnected and rinsed with pure (e.g., A.R.)
acetone (about 15cm^ delivered from a small, all-glass wash bottle), and then
dried at 100-1 10 °C for 3-4 minutes. The electrode with its deposit is weighed
after cooling for about 5 minutes at the laboratory temperature. Cooling in a
desiccator is usually unnecessary and, in any case, requires a much longer time. In
a few cases there is evidence that a small transfer of platinum takes place’ during
electrolysis. If this is suspected, the electrode plus deposit should be weighed, the
deposit dissolved, the electrode weighed again, and the difference taken. The
occurrence of anodic corrosion is most simply detected by weighing the anode
before and after electrolysis.

Electrolysis may be carried out in chloride solutions provided a sufficient

531

Xn, 9 QUANTITATIVE INORGANIC ANALYSIS

amount (1-5 g) of eitliera hyiirazinium ora hydro.\ylainmonium salt {usually the
cliloridc) is added us an anodic depolariser. If no depolariser is added to an acidic
chloride solution, corrosion of the anode occurs and the dissolved platinum is
deposited on the cathode, leading to erroneous results and to destruction of the
anode. A number of metals (for e.xaniple, zinc and bismuth) .should not be
dcpo.->itcd on a platinum surface. These metals, particularly zinc, appear to react
with the platinum in some way, for when they are dis.solved off with nitric acid the
platinum surface is dulled or blackened. Injury to the platinum can be prevented
in these eases by tirst plating it with copper, and then depositing the metal on this
surface; alternatively, a silver gauze electrode may be used. It must be
emphasised that a platinum cathode should have a surface as smooth and bright
as possible, because any surface uncvenne.ss tends to increase during metal
deposition, and may lead U) rough deposits ( Ref. 7).

The determinations described in the following Sections include:
copper (Section XII. 9) : deposition from acid solution :
lead (Section XII, 10); an example of anodic deposition:
cadmium (Section Xll, 1 1 ) and .silver (Section Xll, 12); deposition from cyanide
compic.xes:

copper and nickel in a cupro-nickel alloy (Section XH, 13); an example of
electrolytic separation and of deposition from an aminoniaca! solution;
copper. lead, antimony and (in in a bearing metal (Section XII, 14); an example
of the use of controlled eatlKKic potential.

.XII, 9 COPPER, DLsciiasioih Copper may be deposited from either
sulphuric or nitric acid solution, but. usually, a mixture of the two acids is
employed. If such a solution is electrolysed with an e.m.f. of 2-3 volts the
following reactions occur;

Caihuilc: Cir ’ + 2e Cu

Anode: 40H ~ O, f-4c

The acid concentration of the solution must not be too great, otherwise the
deposition of the copper may be incomplete or the deposit will not adhere
satisfactorily to the cathode. The beneficial effect of nitrate ion is due to its
depolarising action at the cathode:

NOj' + lOfr +3H,0

The reduction potential of the nitrate ion is lower than the discharge potential of
hydrogen, and therefore hydrogen is not liberated in the freestate. The nitric acid
nuist be free from nitrous acid, as the nitrite ion hinders complete deposition and
introduces other complications. The nitrous acid may be removed by boiling the
nitric acid before adding it, or by the addition of urea to the solution:

2H " -i-2NO,- +CO{NlK), - 2N, + COj +311,0

Nitrous acid is most clliciently removed by the addition of a little sulphamic
acid;

ir +N0j"’ + -0-.S0j-NIl. = N, + HS04~ + H,0
the action is rapid, and the acidity of the electrolyte is unaffected. The error due

532

ELECTRO-GRAVIMETRY XH, 9

to nitrous acid is increased by the presence of a large amount of iron; iron is
reduced by the current to the iron(II) state, whereupon the nitric acid is reduced.
This error may be minimised by. the proper regulation of the pH and by the
addition of ammonium nitrate instead of nitric acid, or, best, by the removal of
the iron prior to the electrolysis, or by complexation with phosphate or fluoride.

The solution should be free from the following, which either interfere or lead to
an unsatisfactory deposit : silver, mercury, bismuth, selenium, tellurium, arsenic,
antimony, tin, molybdenum, gold and the, platinum metals,' thiocyanate,
chloride, oxidising agents, such as oxides of nitrogen, or excessive amounts of
iron(III) nitrate or nitric acid. Chloride ion is avoided for two reasons :

1. Chlorine, if set free at the anode, may attack the platinum and some of the
latter may plate out at the cathode: the use of an anodic depolariser, such as a
hydrazinium or hydroxylammonium salt, prevents this.

2. Cu(I) is stabilised as a chloro-complex and remains in solution to be re-
oxidised at the anode.

The electrolytic deposit should be salmon-pink in colour, silky in texture, and
adherent. If it is dark, the presence of foreign elements and/or oxidation is
indicated. Spongy or coarsely crystalline deposits are likely to yield high results;
they arise from the use of too high current densities or improper acidity and
absence of nitrate ion.

Procedure. The solution (lOOcm^) may contain 0.2-0.3 g of Cu (1). Add
cautiously 2 cm^ of concentrated sulphuric acid, 1 cm^ of concentrated nitric acid
(free from nitrous acid by boiling or by the addition of a little urea, or, better,
0.5 g of sulphamic acid), and transfer to, unless already present in, the electrolysis
vessel. For simplicity of description, it will be assumed that the cathode is of the
gauze form (Fig. XII, 8, a) and the anode is a gauze cylinder (Fig. XII, 8, b). Clean
the platinum gauze cathode by heating it in 1 : 1 nitric acid, and washing it
thoroughly with distilled water, followed by pure acetone. Dry the electrode at
100-1 10 °C for 3-4 minutes, cool in the air for about 5 minutes, and weigh.
Handle the electrode by the stem and not by the gauze, since a trace of grease may
cause a non-adherent deposit of copper. Arrange the circuit as shown in Fig. XII,
6, or suitably adapted according to the source of current (if accumulators or
storage batteries supplying not more than 12 volts are employed, one small
resistance will suffice), but do not connect the source of current. Be sure that
the cathode is connected to the negative terminal and the anode to the positive
terminal. If a rotating anode is to be used subsequently, make certain that it can
rotate without coming into contact with the cathode at any point. Place the
electrolysis vessel in position (e.g., a beaker resting on a wooden block or upon a
stand ; a ‘lab-jack’ is very convenient), and adjust the height so that the electrodes
extend nearly to the bottom of the beaker and the cathode is 80-90 per cent
immersed in the solution. Cover the electrolysis vessel (beaker) with a split clock
glass (Fig. XII, 10), and with all the resistance in the circuit, so that only a small
current will flow, close the circuit, and proceed as A or B, depending upon
whether or not stirring is employed.

A. Slow electrolysis, without stirring. With an applied potential of 2-2.5
volts, adjust the resistance until a current of approximately 0.3 amp flows,' as
indicated by the ammeter in the circuit; electrolyse, preferably overnight (2).

mse off the split clock glass and test for complete deposition after the blue
colour of the solution has disappeared. This is best done by adding more water to
raise the level of the electrolyte (say, by 0.5 cm) and continuing the electrolysis for

533

Xll, 9 QUANTITATIVE INORGANIC ANALYSIS

30 minuics. If al the end of this time no copper deposit has appeared on the
freshly immersed surface of the cathode, it may be assumed that Ihedeposiiionis
complete. If a deposit does form, continue the passage of the current as long as
may be judged necessary, and again test as before. If the solution is not to be used
for a further determination, a drop or uvo may be removed and tested with
sodium acetate and hydrogen sulphide water or with potassium
hex.'icyanoferrate(Il) solution.

B, Rapid electrolysis, with stirring. Start the motor driving the anode {or,
less satisfactorily, the glass propeller) and adjust its speed so that the solution is
vigoroiLsly stirred but with no danger of ntcchanical loss of liquid. Use a voltage
aeros.s the terminals of the cell of 3 -4 volts, and adju.st the resistance so that the
current is 3-4 amps. Continue the electrolysis until the blue colour of the solution
has entirely disappeared ( usually somewhat less than 1 hour), reduce the current
to 0.5-1 amp. and test for comp!ciene,ss of deposition by rinsing tlte split clock
glass, raising llic level of tlie liquid by about 0.5 cm by the addition of distilled
water, and continuing the electrolysis for 15 -20 minutes. If no copper plates out
on the fresh surface of the cathode, electrolysis may be regarded .^s complete.

When electrolysis has been shown to be complete, tlie subsequent procedure is
the same whether slow or rapid electrolysis has been employed. Two methods
may be employed : (i) is used only when the residual electrolyte is not required for
further determination, and tii) is of universal application.

(i) Siphon olf the liquid {for example, with a glass siphon provided with a
stopcock). Whilst .liphoning the heavier liquid from the bottom of the beaker,
add water to the top so as to keep the level of the liquid in the electrolysis vessel
nearly constant. Continue this procc.>s until the ammeter needle drops practically
to zero. Then rinse the cathode with A.R. acetone {3), and dry it for 3-4 minutes
at lOO-llO’C. Weigh after cooling in air for about 5minutcs.

{ii) Lower the beaker very slowly, or raise the electrodes, and at the same time
direct a continuous stream of tlisiillcd water from a wash bottle against the upper
edge of the cathode. This washing must be done immediately the cathode is
removed out of the .solution, .md the circuit must not be broken during the
process. When the cathode has been thoroughly washed, break the circuit, dip
the cathode into a beaker of distilled water, and then rinse it with A.R. acetone
(3). Dry at 100-1 iO' C for 3 -4 minutes, and weigh after cooling in air for about 5
minuics.

From the increase in weiglit of t he cathode, calculate the copper content of the
solution. After thecaihode has been weighed, it should becleaned with nitric acid
as described in Section XU, S, and rc-weighed ; the loss in weight will .serve as a
check.

Notes. I . Larger quantiiies of copper may be preseal, particularly if rapid
electrolysis is employed; the quantity given is, however, convenient for
instructional purposes'. For practice in the determination, prepare the solution
eit/icr by weighing out accurately about 1.0 g of A.R. copper sulphate
peniahydrate or by di.ssolving about 0.25 g, accurately weighed, of A.R, copperin
1 : 1 nitric acid, boiling to remove nitrous fumes, just neutralising with ammonia
solution, and then just acidifying with dilute sulphuric acid and diluting to 100
cm-*.

2. If conditions do not allow' this practice, use a current of 1. 5-3.0 amps:
deposition is usually complete in 2-4 hours. This procedure is, Itowevcr, less
satisfactory ifstirring is not employed.

534

ELECTRO-GRAVIMETRY XH, 10

3. This is best effected by directing a stream of A.R. acetone from a small all-
glass wash bottle on to the electrodes : about 10-15 cni^ are required. •

XII, 10. LEAD. . Discussion. Lead is deposited, quantitatively as the dioxide
at the anode in the presence of a high concentration of nitric acid (10-15 cm^ of
concentrated acid per 100 cm^ of the electrolyte). The addition. of 3-4 drops of
concentrated sulphuric acid is said to make the deposit more adherent. It is
probable that the lead is oxidised to the Pb(IV) state in the nitric acid medium,
and the Pb(IV) ion is converted into hydrated lead dioxide, Pb02,xH20, at the
anode. Alternatively, the Pb(Ilj ion is oxidised quantitatively to hydrated lead
dioxide at the anode. The action of nitric acid is an example of a cathodic
depolariser. Nitrate ion is reduced more easily than Pb(II) ion, and thus
functions as a cathodic depolariser to maintain the cathode potential below the
value required for the reduction of the lead ion. With a gauze electrode of the
usual size about 0. 1 g of lead is the maximum that can be firmly deposited.

It is difficult to remove all the water from the electrode by drying at low
temperatures. For a temperature of 120 °C, a conversion factor of 0.864, instead
of the theoretical conversion factor of 0.8662, is employed. A useful method is to
dissolve the dioxide, without drying, in standard oxalic acid solution, and to
titrate the excess acid with standard potassium permanganate solution.

The following interfere in this determination : mercury, arsenic, antimony, tin,
selenium, tellurium, phosphorus, chromium, chloride, iodide, silver, bismuth,
and manganese (the last three metals tend to form peroxides at the anode).

Procedure. For a platinum gauze electrode, the solution ( 1 00 cm^) should
contain not more than 0.1 g of Pb as lead nitrate, 15 cm^ of concentrated nitric
acid (free from nitrous acid), and none of the interfering elements mentioned
above.

Heat the anode to 1 20 °C in an electric oven for 20-30 minutes, allow to cool in
air for about 5 minutes, and weigh. Connect the positive terminal of the source of
current to the gauze anode, and the negative terminal to the wire or gauze
cathode (1). Adjust the current to 0.05-0.1 amp at 2 volts with the aid of the
rheostat, and allow the electrolysis to proceed overnight. Test for completeness
of deposition by adding about 20 cm^ of distilled water and continuing the
electrolysis for 15 minutes; if no darkening of the freshly covered anode surface
occurs, deposition is complete. When the electrolysis is complete, either lower the
beaker from the electrodes or slowly raise the electrodes out of the solution
without interrupting the current, and at the same time rinse the electrodes very
thoroughly with a jet of water from a wash bottle. Then disconnect the source of

current, wash the anode with A.R. acetone, and dry at 120 °C for 20-30 minutes;
cool in air for about 5 minutes and weigh.

The determination may be carried out more rapidly by one of the following
methods. In these the outer electrode is the anode.

(a) With a current of 1. 5-2.0 amps at 2 volts at the ordinary temperature.

Electrolysis is complete in about 1 .5 hours.

{b) With 2 volts and an initial current of 0.5 amp, which is subsequently raised to
5 amps. 0. 1 gram of lead dioxide may be deposited in 6-8 minutes at room
temperature : this time is still further reduced by working at about 60 °C.

(c) Up to 0.3 g of lead may be deposited in about 10 minutes from a solution
(total volume 85-100 cm^) containing 10 cm^ concentrated nitric acid by
electrolysis at 90-95 °C, with a current of 5 amps. Under these conditions

535

XU, 11 QUANTITATIVE INORGANIC ANALYSIS

clcctro-o-sniosis expels most of the water from the deposit, and it may be

dried by washing with A.R. acetone in the ordinary svay.

The deposited dioxide is removed from the electrode, after weighing, by
immersion in warm 1 : 1 nitric acid to which a little pure hydrogen peroxide !i;is
been added.

Note. I . Full experimental details of the general technique are given under
Copper (Section Xll, 9), and will therefore not be repeated in the description of
this and the succeeding determinations.

XII, 11. CADMIUM. Dhcussioit. Cadmium is Itesl determined from a
faintly alkaline .solution containing only enough potassium cyanide to keep the
cadmium in solution, i.e., containing the complex K^ICdlCNl^j {Procalure .d).
Elements, .such as zinc and silver, interfere, l.ess accurate results, but sufficiently
precise for most routine analyses, may be obtained in very dilute sulphuric acid
.solution {Praci'ilure P). The elements that interfere are essentially those
mentioned under Copper (Section .Xlf, 9). Deposition may al.so be made from a
hydrochloric acid solution in the presence of hydroxylaminonium chloride or
hydrazinium chloride, which acts us an anodic depolari.ser{/'/-oa‘f/:/ri' C).

Procedure -d. The solution should contain about 0.4g of Cd as the
sulphate, acetate, or, less desirably, the nitrate. Add a drop of phcnolphthalcin,
followed by cu. O.LU-sodium or pota.s^ium hydroxide until a permanent pink
colour is just obtained. Then add a solution of pure (e.g„ A.R.) potassium or
sodium cyanide dropsvise and with constant stirring until the precipitated
cadmium hydroxide ju.st di.s.solves. A large c.xces.s of alkali cyanide should be
avoided. Dilute to l()d-I50cm-’, and electrolyse tiie cold solution, preferably
with a platinum gauze cathode, and a current of 0..s -0.7 amp at 4.il-5.0 volts ( D.
At the end of 6 hours, increase the current to 1,0-1. 2 amps, and continue the
elcxtrolysi-s for anotlver hour. Wash the split clock gla.ss and the sides of the
beaker with about 20cm'' of water, and continue the electrolysis for 15 minutes,
if the newly exposed surface of the cathode remains bright, thus indicating that
the deposition of the cadmium is complete (2), remove the electrolyte from the
electrodes, rinse immediately wiili water, .slop the current, and rinse the cathode
with ethanol or A.R. acetone. Dry at 100 C, cool, and weigh. Test the residual
electrolyte for cadmium by any of the recognised tests.

The determination may be carried out more rapidly by using a rotating
cathode with a current of 1 .5-2.0 amps at 2.7-3,0 volts. 0.2 gram of Cd may thus
be deposited in 30 minutes.

Procedure B. The cadmium should be present in the solution (lOOcm^) as
sulphate; nitrates and chlorides must be absent. The maximum concentration of
free sulphuric acid is O.SiV, and 5 g of poi;issium hydrogcnsulphate is added.
Electrolyse :it room temperature with 0.1- 0.2 amp at '2.4-2.8 volts; after 3 hours
increase to 0.5 amp until electroly-sis is complete. 0.3 gram of Cd is thus deposited
in 3-4 hours.

Alternatively, use a rotating electrode with I.5“1.7ampsat 2.7 volts. 0,2 gram
of Cd may thus be depo.sited in 20 niinute,s.

Procedure C. The cadmium (up to 0.3 g) should be present in the solution
(200 cm^) as chloride. Add 2g of hydro.xylammonium chloride or hydrazinium
chloride, acidify slightly with hydrochloric acid, and electrolyse at the laboratory
temperature, using a current of about 1 amp.

Notes. 1. If left overnight, use a current of0.2--0.3 amp at 2.8-3.2 volts.

536

ELECTRO-GRAVIMETRY XD, 12/13

2. It is sometimes difficult to detect the deposition of the bright eadmium on
the platinum surface. This difficulty is readily overcome by heavily plating the
platinum electrode with copper or silver first and then proceeding with the
electrolysis in the usual manner. An added advantage is that the removal of
the cadmium after the electrolysis is easier; also, if the temperature of drying
should accidentlly exceed 100 °C, there is little danger of harming the
platinum electrode. >

Xn, 12. SILVER. Discussion. Silver may be determined by electrolysis in
nitrate, ammoniacal, or cyanide solutions. In cyanide solution the silver is
present largely as the complex ion :

[Ag(CN)2]-^Ag++2CN-

an excellent plate is obtained, and separation from other elements (e.g., copper
and lead) may be effected. The cyanide method will be described.

The disadvantage of the electrolytic method is that so many other elements are
also deposited, either wholly or in part, that a number of preliminary separations
are usually required before it can be applied. For this reason,, it is not widely
employed.

Procedure. The silver (ca. 0.2 g) should be present in neutral or faintly
acidic solution as the nitrate. Add pure (e.g., A.R.) potassium cyanide until the
precipitate of silver cyanide is dissolved, and then add an excess such that, about
2g of potassium cyanide is present in the solution. Dilute to 100-120cm^.
Electrolyse with 0.2-0.5 amp at 3. 7-4.8 volts at 20-30 °C; about 0.1 g of Ag is
deposited in 3 hours. Alternatively, electrolyse with a rotating electrode with
0.5-1.0 amp. at 2.5-3.2 volts; 0.2 g of Ag is deposited in 20-25 minutes.
Completeness of deposition is tested for by transferring a few drops of the
electrolyte to a test-tube, acidifying with a little nitric acid, boiling off the
hydrocyanic acid (caution: poisonous), rendering ammoniacal, and adding a few
drops of ammonium sulphide solution ; no brown precipitate should be obtained.
The determination is completed as under Copper (Section XII, 9).

Note, If insoluble silver salts are to be analysed, e.g., chloride, bromide,
iodide, and oxalate, these may be dissolved directly in the potassium cyanide
solution.

xn, 13. ELECTROLYTIC SEPARATION AND DETERMINATION OF
COPPER AND NICKEL. Discussion. This determination has been included
to indicate the use of constant current electrolysis in the separation and
detennination of metals in simple alloys. More complex alloys require the
application of methods utilising controlled potential at the cathode ; see Section
XII, 14. The theory of simple separations is discussed in Section XD, 5. There are
a number of alloys, which include Monel metal, certain coinage alloys, and
cupro-nickel,’ which are composed principally of copper and nickel, together
with small amounts of iron and manganese and not more than traces of other
elements. These are suitable for electrolytic separation.

The copper is determined in strongly acid solution at a potential not exceeding
4 volts (above this potential nickel may plate out). The solution is evaporated to
turning in order to remove excess of nitric acid, the iron present is precipitated
"hh- solution, and the nickel deposited from the filtrate after the

addition of a large excess of ammonia solution.

537 ,

XII, 13 QUANTITATIVE INORGANIC ANALYSIS

Procedure {analysis of a copper- nickel alloy). Weigh out accurately about
0.5 g of the clean alloy into a 1 50-cm^ tall form beaker, which should be suitable
as an electrolytic vc.ssel. Add a rni.xture of lOcni^ water, 1 cm-* concentrated
sulphuric acid, and 2 cm-* concentrated nitric acid to dissolve the alloy. When
solution is coniplelc, boil off the oxides of nitrogen, and dilute to lOOcm^, The
solution is now ready for the deposition of copper.

Copper. Proceed as directed in Section XII, 9. employing either the slow or
rapid method of electrolysis. Wash the copper deposit thoroughly with water,
and keep the solution for the determination of iron and nickel.

Iron. Evaporate the solution and washings from which the copper has
been removed on a low-temperat are hot plate a.s far as po.ssib!c, and then heat at a
higher temperature until fumes of sulphur trio.xide appear. Cool the residue, and
carefully add water until the volume is about 25 cm^. Precipitate the small
quantity of iron that is now present in the Fc(lJD state by adding to the warm
solution about lOcm^ of 1 1 1 ammonia solution in e.xccs.s. Filter through a small
quantitative filter paper, and collect the filtrate in a 15()-cm^ electrolysis beaker
(A). Wash the preeipitafe three times with water. Place the original beaker under
the filter; dis.solve the precipitate in a little hot 1 1 5 sulphuric acid and wash the
paper with water. Precipitate the iron ag.iin with the same large e.xo.'ss of K!
ammonia solution, and filter through the same paper. Wash the precipitate, and
collect the filtrate and washing.s in beaker (.-Dconlaining the filtrate and washings
from the first precipitation. Ignite and weigh as Fe,Oj (Section XI, 21).

Alternatively, since the iron content is small, the washed precipitate may be
dissolved in dilute hydrochloric acid and the iron determined coloriinelrically
(Section .XVIII, 21).

Nickel. Add 15em^ conecmraied ammonia solution to the ainmoniacai
nickel solution, and dilute to 100- 120 cm^. Carry out electrolysis using a rotating
electrode with a current of 4 amps at .3- 4 volts; 0,1 g of nickel is deposited in
about 10 minutes. Tc.st for completeness of precipitation in the usual way by
adding about 20 cm^ of water, and continuing the electrolysis for 15-20 minutes;
no nickel sliould be deposited on the freshly immersed surface. Alternatively, the
diincthylglyo.ximc test may be applied after neutralisation of the ammoniacal
solution with hydrochloric acid. Wash, dry, and weigh the cathode as in the
determination of copper.

The nickel is best remosed from the electrode by means of dilute nitric or
sulphuric acid; concentrated nitric ;icid should not be employed because of the
danger of inducing passivity. If dilficuliy is experienced in stripping the nickel
from the platinum cithode, anodic solution of the metal in warm dilute nitric
acid may be employed.

Electrolytic determinations with controlled cathode
potential

The principles of electrolysis using controlled cathode potentials have been
discussed in Section XII, 7, and the dclail.s for determination of .anlimony,
copper, lead, and tin in a bearing metal, which are given below, serve to illustrate
the practical details of this procedure. As indicated previously, the cathode
potential may be controlled manually, but it is preferable to make use of

538

ELECTRO-GRAVIMETRY XH, 14

commercially available equipment, a potentiostat, which will automatically
maintain the potential of the working electrode constant.

Xn, 14. ANTIMONY, COPPER, LEAD, AND TIN IN AN ALLOY (e.g.,
BEARING METAL). Weigh accurately 0.2-0.4 g of the alloy (as drillings or
fine filings) into a small beaker. Dissolve the alloy by warming with a mixture of
10 cm^ concentrated hydrochloric acid, lOcm^ water, and Ig ammonium
chloride (the last-named to minimise the loss of tin as tetrachloride). Solution
may be hastened by the addition, drop by drop, of a saturated solution of
potassium chlorate or of concentrated nitric acid. When all the alloy has
dissolved, boil off the excess of chlorine or of nitrous fumes, add 5cm^
concentrated hydrochloric acid, dilute to 150cm^, and then add 1 g hydrazinium
chloride. Stir the solution efficiently and electrolyse, limiting the cathode
potential to -0.36 volt vs. S.C.E. ; copper and antimony are deposited together.
After 30-45 minutes the current becomes constant (usually at about 20
milliamps) : remove the saturated calomel electrode, stop the stirrer, lower the
electrolysis beaker, and at the same time wash the electrodes with a fine stream of
water from a wash bottle directed at the upper rims. Now break the circuit,
remove the cathode, rinse it with A. R. acetone, dry for 3-4 minutes at 1 05 ° C, and
weigh after cooling in air for 5 minutes.

Separate the copper and antimony by dissolving the deposit in a mixture of
5 cm^ concentrated nitric acid, 5 cm^ 40 per cent hydrofluoric acid (CARE), and
lOcm^ water: boil off the oxides of nitrogen, dilute to 150cm^, and add dropwise
a solution of potassium dichromate until the liquid is distinctly yellow. Deposit
the copper by electrolysing the solution at room temperature and limiting the
cathode-S.C.E. potential to —0.36 volt. Evaluate the weight of antimony by
difference.

To the solution from which the copper and antimony have been separated as
above, add 5 cm^ concentrated hydrochloric acid and 1 g hydrazinium chloride.
Electrolyse, using a copper-plated cathode ;* the purpose of this is to prevent
alloy formation of the platinum with the lead. Add water to the solution until the
cathode is completely immersed, and then electrolyse with the cathode
maintained at —0.70 volt V5. S.C.E. Continue the electrolysis for 45 minutes: the
final value of the current is, in this instance, often an unreliable indication of the
completeness of the deposition. Neutralise the electrolyte by adding dilute
ammonia solution (1 '. 1) — otherwise the deposited metals will partially dissolve
during the washing process — immediately lower the electrolysis beaker, wash the
electrodes with water, rinse the cathode with A.R. acetone, dry, and weigh in
the usual manner. The increase in weight gives the weight of lead and tin in
the sample.

fhe deposit from the cathode in 15 cm^ nitric acid, sp. gr. 1.20, in a
400-cm beaker, and finally wash the cathode with water. Evaporate the resulting
^lution almost to dryness, cool, and add a further 15 cm^ nitric acid, sp. gr. 1.2.

igest hot for a time and then filter the hydrated tin(IV) oxide on a paper-pulp

Prepare this electrode by plating about 50 mg of copper — use a measured volume, e.g., 25.00 cm^ of
a standard copper sulphate solution, say, 0.00500A/-from a H^SO^-HNOj solution, washing

wa er, tollowed by A.R. acetone, drying at 1 10 °C for 3—4 minutes, and weighing after cooling
lor D minutes m air. °

539

XU, J5/16 QUANTiTATlVElNORGAN'ICANALYSIS

pack, and wash it four limes with hot water. Dilute the resulting filtrate and
washings to 100cni\ and heat to boiling, l-leclroly.se the hot solution with a
small platinum gauze anode at +“5 amp.s until the deposition of PbO^ is complete
(about 5 minutes), Remove the anode, dry. and weigh a.s before. Calculate the
percentage of lead from the weight of PbO^ using the empirical factor of 0.864.
Evaluate the tin content by subtraction from the combined weight of tin and lead.

Calculate the percentages of antimony, copper, lead, and tin in the alloy.

XII, 15. INTERN.XL ELECi'UOLYSIS. The term internal electrolysis was
applied by H. J. S. Sand (1930) toelectro-analy.'>i-s in which an attackable anode is
used and there is an e.xternal wire connection between the cathode and anode so
that electrolysis proceeds spontaneously without the application of an e.xternal
e.ni.f. The arrangement is, in eticet. a short-circuited voltaic cell. Internal
electrolysis has al .%0 been described a.s .spontaneous clectro-gravimctric analysis.
The method is a special case of controlled potential electro-analysis using a
platinum gauze cathode: potential control is achieved by appropriate choice of
anode, and no external voltage source is required. The driving voltage is, of
course, small and. in eonsequcace. the celt resistance is a critical factor in
determining the rale of metal deposition. The applicalion.s of the procedure are,
in general, restricted to the determination of small amounts ( ?*35 mg) if the lime
of electrolysis is not to be excessively long. See Refs, 8 and 9,

Applications of interna! electrolysis arc mainly eontined to the determination
of small amounts of relatively noble metal impurities in relatively base metals or
alloys. These include:

Silver in lead, galena, and pyrites.

Mercury in copper and brass.

Copper in lead and m steel.

Bismuth and copper in lead, in lead ■ tin alloys, and in galena.

Lead and also cadmium in zinc.

XU, 16. Rcrcrcnccs

1. fl, Dichl (1948). iiltrirodumicul .Ui.ilyMs ntili OVui/ci/ CoiIuhIi.’ hwiiiuil Coiiiral.
Colunibu.s, Ohio; G. 1-. Smil)i Chemical Co.

2. C. W. C, .Nfilner .md R. N. Whiitcm 1 1'>52>. Coninilh'd Poiaiiiiit in the Amlyiis of
Cuppcr-haur Alloys. .Aiiulysl, 77. 11.

3. B. Alfoosi (19.38). 'Deicriniiialioii of Copier. Lead. Tin and .Aiuiinouy by Conirolled
I’oicnlial Electrolysis". Parts 1-111. .In, Owmea Acui, 19. 276; 389; .369.

4. J. 1- fferingshaw and P. F, llallludc ( I960). 'A PoteiiUosi.n lor Electrograviinetrie
Analysis' Aiuitysi, 85. 69.

5. J. A. Maxwell and R. P. Graham (1950). ‘TIicMcrcutv Cathode and its .Xpplications',
C/u7«. /Zei'.,46,471.

6. J. A. Page, J. A. Maxwell and R, P Graham (1962). ■.Anaiyiieal Ap()lications ot the
Mercury Electrode’. Aiiolyd, 87, 245

7. II. J. S. S.ind (1940). Ekxtruclu’mi.urv and Ekrinn-lniiwat Amdysis. Vols. I and 11.
Blaekio: London.

8. B. L. Clarke, L. A. Woolen and C. L. Lu);c (1936). "Analysis by "Internal”
Electrolysis', /mil. Eny. Ciwnt.. Amd. Edn.. 8. 4 1 1 .

9. B, L. Clarkeand L. A. Wooten ( 1939). ‘Imcrnai Electrolysis asa Method of Analysis".
Trans, Ehrtrochem. Soc., 76, 63,

540

ELECTRO-GRAVIMETRY XH, 17

Xn, 17. Selected Bibliography

1. D. R. Browning (1969). Electrometric Methods. London; McGrawHill.

2. C. W. Davis (1967). Electrochemistry. London; Geo. Newnes Ltd.

3. P. Delahay (1957). Instrumental Analysis. New York; Macmillan Co.

4. G. A. Ewing (1975). Instrumental Methods of Chemical Analysis. 4th edn. New York;
McGraw Hill.

5. 1. M. KolthoffandP.J.Elving(1959). Treatise on Analytical Chemistry. Partl.Yol. 4.
New York; Wiley.

6. H. A. Laitinen (1960). Chemical Analysis. An Advanced Text and Reference. New
York ; McGraw Hill.

7. J. J. Lingane (1958). Electroanalytical Chemistry. 2nd edn. New York; Interscience.

8. G. W. C. Milner (1957). The Principles and Applications of Polarography and other

E/ec/roana/yt/ca/ /’rocme5. London; Longmans Green and Co. ' ' '•

9. W. F. Piekering (1971). Modern Analytical Chemistry. New York; Marcel Dekker
Inc.

10. H. A. Strobel (1960). Chemical Instrumentation. A Systematic Approach to
Instrumental Analysis. Reading, Mass; Addisoii-Wesley.

11. C. R. N. Strouts, J. H. Gilfillan, and H. N. Wilson (1962). Analytical Chemistry. The
Working Tools. Vol. 1. 2nd edn. London; Oxford University Press.

12. F. J. Welcher (1966). Standard Methods of Chemical Analysis. Vol. 3-A. 6th edn.

Princeton; Van Nostrand. ■ ' ■

13. H. H. Willard, L. L. Merritt, and J. A. Dean (1974). Instrumental Methods of Analysis.
5th edn. New York; Van Nostrand.

14. C. L. Wilson and D. W. Wilson (1964). Comprehensive Analytical Chemistry. Part A.
Vol. 2. Amsterdam ; Elsevier.

541

xn. 15/16 QUANTITATIVE INORGANIC ANALYSIS

pack, and wasli ii four timc.s with hoi water. Dilute the resulting filtrate and
washings to lOOcm^ and heat to boiling. I-lcctroly.se ihe hot solution with a
small platinum gauze anode at 4-5 amp.s until the deposition of PbOj is complete
(about 5 minutes). Remote the anode, dry, and weigh a.s before. Calculate the
percentage of lead from the weight of PbO. using the empirical factor of 0.86-},
Evaluatuthc tin content by .subtraction from the combined weight of tin and lead.

Calculate the percentages of antimony, copper, lead, and tin in the alloy.

,XI1, 15. INTERN.AL El.Ed ROLYSIS. The term internal electrolysis was
applied by H. J. S. Sand ( 1930) to clectro-analysisin which an attackable anode is
vised and there is an c.xternal wire connection between the cathode and anode so
that electrolysi.s proceeds spontaneously without the application of an e.xternal
c.m.f. The arrangement is. in elfect. a short-circuited voltaic ceil. Internal
electrolysis has also been described as spontaneous electro-gravimetric analysis.
The method is a special case of controlled potential clcctro-analy.sis using a
platinum gauze cathode: potential control is achieved by appropriate choice of
anode, and no c.xternal voltage source is rcquirctl. The driving voltage is, of
course, small and, in consequence, the cell rc-Nistance is a critical factor in
determining the rate of metal tieposition. The applications of the procedure are,
m general, restricted to the determination of small amounts ( >25 mg) if the time
of electrolysis is not to be excessively long. See Refs. S and 9.

Applicalioiu of interna! electrolysis arc mainly confined to the determination
of small amounts of relatively noble metal impurities in relatively base metals or
alloys. These include:

Silver in lead, galena, and pyrites.

Mercury in copper and bra.ss.

Copper in lead and in steel.

Uismuth and copper in lead, in lead -tin alloys, and in galena.

Lead and also cadmium iu zinc.

XII, 16. References

I 11, Dichl (19-lS). Eli’cirai lic/iuail Amily.^h hiili GrtiucJ Ctiihoih' PoU-ttiitil Ciinlrol.
Columbu.s, Oliio; G. I-, Smith Chemical Co.

2. C. \V. C. Milner and R. N. Wlmietn (1952). CoiunilU ii I'otauiaJ in f/ie .t/w/nd a/
CoppiT-luisc Alloys. Analysi, 77. 1 1.

3. It. Altonsi (1958). ‘Deierminaiion of Copper, l.cail, Tin .ind Antimony by Coniroilcd
Poteniial HleetrolysiC. Parts 1~1II. An. Clwiiscn .drJu. 19. 276; 3S9; 569.

4. J. 1-. Heringshaw .ind P, 1-. Hallliide (I960). *A Poientio.slat for Electrogravimetric
Anaiysw’. Aiiolysl, 85. 69.

5. J. A. Mu-xwel! and R. P, Graiwm (1950). 'TIk Mercury Cathode and iis.-Xpplieations'.
Clu-ni. Rcv.,A6. -171.

6. J. A. Page, J. A. Maxwell and R. P. Graham (1962). 'Analytical Applications of ihe
Mercury Electrode’. Analyn, 87, 245

7. II. J, S. Sand (19-10). Elcarocln'mistry anJ likcirocliankal Aimlysis. Vols. I and II.
Blackie: London.

8. U. L, Clarke, L, A. Wooten and C, L. Luke (19.36). 'Analysis by “Internal"
Electroly.sis’. /mil. I-Mg. Cltait., .4mil. Eilm, 8, 41 i .

9. B. L. Clarke and L. A. Wooten { 1 939). 'Internal Electrolysis a.s a Method of Analysis’.
Trans. Elecirochcm. Stic., 76, 63.

540

ELECTRO-GRAVIMETRY XH, 17

Xn, 17. Selected Bibliography

1. D. R. Browning (1969). Electrometric Methods. London; McGraw Hill.

2. C. W. Davis (1967). Electrochemistry. London; Geo. Newnes Ltd.

3. P. Delahay (1957). Instrumental Analysis. New York; Macmillan Co.

4. G. A. Ewing (1975). Instrumental Methods of Chemical Analysis. 4th edn. New York;
McGraw Hill.

5. I. M. Kolthoff and P. J. Elving (1959). Treatise on Analytical Chemistry. Part I. Vol. 4.
New York; Wiley.

6. H. A. Laitinen (1960). Chemical Analysis. An Advanced Text and Reference. New
York; McGraw Hill.

7. J. J. Lingane (1958). Electroanalytical Chemistry. 2nd edn. New York; Interscience.

8. G. W. C. Milner (1957). The Principles and Applications of Polarography and other

E/ectroana/yt/ca/ EroceMCi. London; Longmans Green and Co. '

9. W. F. Piekering (1971). Modern Analytical Chemistry. New York; Marcel Dekker
Inc.

10. H. A. Strobel (1960). Chemical Instrumentation. A Systematic Approach to
Instrumental Analysis. Reading, Mass; Addisori-Wesley.

11. C. R. N. Strouts, J. H. Gilfillan, and H. N. Wilson (1962). Analytical Chemistry. The
Working Tools. Vol. 1. 2nd edn. London; Oxford University Press.

12. F. J. Welcher (1966). Standard Methods of Chemical Analysis. Vol. 3-A. 6th edn.
Princeton; Van Nostrand.

13. H. H. Willard, L. L. Merritt, and J. A. Dean (1974). Instrumental Methods of Analysis.
5th edn. New York; Van Nostrand.

14. C. L. Wilson and D. W. Wilson (1964). Comprehensive Analytical Chemistry. Part A.
Vol. 2. Amsterdam; Elsevier.

541

CH^(\PTER Xtll COULOIVIETRY

XIII, I. GENERAL DISCUSSION. Coulotnciric analysis is an application
of Faraday’s first law of elect roly sis which may be expressed in the form that the
extent of chemical reaction at an electrode is directly proportional tothequantity
of electricity passing through the electrode. For each equivalent of chemical
change at an electrode 964S7 coulombs of electricity (the Faraday constant) are
required; a coulomb is that quantity ofelectricity represented by the flow of one
ampere for one second.

The fundamental requirement of a coulometric analy sis is that the electrode
reaction used for the determination priKccds with 100 percent efliciency so that
the quantity of substance reacted can be c.xpressed by mean.s of Faraday's law
from the measured quantity of electricity (coulombs) passed. The substance
being determined may directly undergo reliction at one of the electrodes {primary
couhmctric amily.ui), or it may react in solution with another substance
generated by an electrode reaction {.sca»ultiry cotihiinctric aiuilyxi.t).

The weight corresponding to one equivalent of substance being electrolysed is
its atomic weight or its molecular weight divideti by the number of electrons
involved in the electrode reaction. The weight IF of substance produced or
consumed in an electrolysis involving Q coulombs is therefore given by the
c.xpression

IF.

96487//

where IF„ is the atomic weight or the molecular weight of the substance being
electrolysed, and n is the number of electrons involved in the electrode reaction.
Analytical methods based upon the measurcineiu of a quantity ofelectricity and
the application of the above equation are termed coulometric methods— a tenn
derived from ‘coulomb*.

Two distinctly dilTcrent coulometric techniques arc available: (i) coulometric
analysis with controlled potential of the working electrode, and (ii) coulometric
analysis with constant current. In the former method the substance being
determined reacts with 100 per cent current efliciency at a working electrode the
potential of which is controlled. The completion of the reaction is indicated by
the current decreasing to practically zero, and the quantity of the substance
reacted is computed from the reading of a coulometcr in series with the cell or by
means ol a current-time integrating device. In method (ii) a solution of the
substance to be determined is electrolysed with constant current until the

542

COULOMETRY XIII, 2

reaction is completed (as detected by a visual indicator in the solution or by
amperometric, potentiometric, or spectrophotometric methods) and the circuit is
then opened. The total quantity of electricity passed is derived from the product
current (amperes) x time (seconds) : an accurate electric stop-clock may be used
or, more conveniently, a low inertia integrating motor-counter unit.

Xni, 2. COULOMETRY AT CONTROLLED POTENTIAL. In a con-
trolled potential coulometric analysis, the current generally decreases
exponentially with time according to the equation

/, = 7oe“'‘'‘ or' /, = /olO-'"

where /q is the initial current, /, the current at time t, and k {k') is a constant. It can
be shown (Lingane, Ref. 7) that A: = 25.8DA/(5P, where D is the diffusion
coefficient of the reducible substance (cm^ s“ ^), A is the electrode area (cm^), 3 is
the thickness of the diffusion layer (cm), and V is the total volume (cm^) of the
solution of concentration C. A typical time-current curve is shown in Fig. XIII,
1, the current decreases more or less exponentially to almost zero. In many cases
an appreciable ‘background current’ is observed with the supporting electrolyte
alone, and in such instances the current finally decays to the background current
rather than to zero ; a correction can be applied by assuming that the background
current is constant during the electrolysis. The reaction, strictly speaking, is never
complete; nevertheless, when the ratio /,/Io reaches a sufficiently low value (e.g.,

0.00 1 ), the analysis may be terminated.

In electrolysis at controlled potential, the quantity of electricity Q (coulombs)
passed from the beginning of the determination to time t is given by ■
r<

Q =

Itdt

■0

where I, is the current at time t. ’

The integration may be performed graphically by measuring the area under the
current-time curve or automatically by means of a mechanical current-time
integrator. Alternatively, I, may be measured at a series of suitable time intervals,
and then log/, plotted against f, a straight line of slope equal to k'/2.303 is
obtained. Then

Q = j I,.dt = \ I Q.e '" ‘.dt = IJk’ for large values of t.

Jo Jo ■

Clearly, this process is time-consuming and it is better to use a coulometer or an
integrating device.

The apparatus employed in controlled potential coulometry , may. be
considered under three headings ;

1 . the coulometer or other method for determining the quantity of electricity ;

2. the controlled source of current;

3. the electrolysis vessel.

Coulometers suitable for measuring the total quantity of electricity passed
include (i) the silver coulometer, (ii) the iodine coulometer, (iii) the
hydrogen-oxygen coulometer, and (iv) the hydrogen-nitrogen coulometer ; the
coulometer is connected in series with the electrolysis vessel. Of those listed, the
silver coulometer is the most accurate, and consists of a platinum basin which
serves as cathode for the electrolysis of 10 per cent silver nitrate solution, together

543

Xni,2 QUANTITATIVE INORGANIC ANALYSIS

with a silver rod anode. A small filter cnicible,
also containing silver nitrate solution is placed
inside the basin, and the silver anode b
situated within the crucible; in this way, any
small particles which may break off from the
anode are prevented from adhering to the
basin, the increase in weight of which
me;isures the quantity of electricity passed.

The iodine coulomeler contains a pair of
platinum electrodes immersed in potassium
iodide solution; at the end of the de-
termination, the liberated iodine is titrated
with standard thiosulphate solution and the
number ofcoulombs passed can be calculated.

Trac. t The hydrogen-o.xygen coulometer consists of
a tube about -JO cm in length and 2 cm interna!
Fig. Xlll, 1 diameter, terminating in a tap at its upper end

and with the lower end joined by Ilcxiblc tubing to a levelling tube. Two pUitinum
electrodes are sealed into the lower end of the tube, and the upper end, which is
graduated, is surrounded by a water jacket so that ga.s within the tube can be
maintained at const;int temperature. The electrolyte is 0,5.1/ pot;issium sulphate
solution which must be saturated with hydrogen and o.xygcn just prior to the
e,\pcriment ; this is done by passing ;i current of 50-100 mA through the solution
for about 5 minule,s with the tap open and the liquid close to the top of the tube.
The .stop-cock is then closed, the liquid levels in the two tubes equalised, and then
the coulometer is attached to the m:tin electrolysis circuit and the current
switched on: the temperature and the barometric pressure must be recorded.
When the volume of gas no longer incretises, ns volume is read and corrected to
s.t.p. {vapour pressure of the pota,ssium sulphate solution I7.2mm (20');
23.2mm (25); 3l,2mm (30')}. ThcoTciic;jlly, the volume of the
hydrogen-o.xygen mi.\ture at .s.t.p. .should be 0.1741 enf* per coulomb, but
the observed value i.s 0.1739 cm'*. On account of this discrepancy, the
hydrogen-nitrogen coulomeler is often preferred; iliis is set up and used ;is
described above, but the electrolyte is u solution of hydnuinium sulphate.

A method sometimes u.scd to measure the quantity of electricity passed is to
include a standard resistor in the circuit and to connect a polcntiometric recorder
across the resistor. Upon completion of the electrolysis, the chart below the
recorder trace is cut out and weighed on an analytical balance, thius permitting
evaluation of the time-current integral. For accurate work however this method
is unsuitable owing to variations in the density of the paper, ;ind Bishop (Ref 1)
has described an accurate procedure based upon resistance-capacity integration.

The current source for the electrolysis may be a large storage battery or a mains
operated power-supply unit together with a large series resistor. A simple circuit
showing how the electrode poieniial may be controlled manually is shown in Fig.
Xlll, 2 (compare Fig. XU, 3), but the coulometer or other device for measuring
the quantity of electricity is not included. The ammeter A indicates the
electrolysis current : the voltmeter V records the e.m.f. being applied between the
anode and the cathode (total applied voltage). The potential of the cathode with
respect to the reference electrode S.C.E. (usually a saturated calomel electrode) is
directly indicated by a high-resistance voltmeter G previously calibrated against

544

COULOMETRY XIU,2

a potentiometer* or, better, a pH meter
provided with a millivolt scale. The
experimental procedure consists in ad-
justing the resistance manually until the
potential difference between the cath-
ode and the reference electrode attains
the desired value. As the electrolysis
proceeds, the cathode tends to become
more negative with respect to the refer-
ence electrode, and it is necessary to
adjust the rheostat so as to restore the
cathode to the desired potential. The
ammeter reading decreases throughout
the electrolysis and generally attains a
low constant value signalling the com-
pletion of the determination. Frequent
adjustment and constant attention are demanded by this procedure if the cathode
potential is to be kept constant to ±0.05 volt or better. The development of
instruments, known as ‘potentiostats’, which automatically maintain the
potential of an electrode constant to ±5-10 millivolts at any predetermined
value, has led to considerable development of the method. A number of
potentiostats have been described in the literature (Refs. 2, 3), and many
instruments are commercially available : some of these (e.g., Solartron, Beckman
‘Electroscan’, McKee-Pedersen), can function as both potentiostats and
amperostats (producing constant current).

The electrode whose potential is being controlled (which may be either the
cathode or the anode) is generally called the ‘working electrode’ of the cell. The
non-controlled electrolysis electrode is termed the ‘auxiliary electrode’, and the
third electrode is a ‘reference electrode’; this does not conduct any of the
electrolysis current and merely serves to permit observation of the potential of
the working electrode.

Two types of electrolytic cell, due to Lingane, suitable for coulometric analysis
at controlled potential will be described ; both use a mercury cathode. In the first
(Fig. XIII, 3) the cell has a capacity of about 1 00 cm^ and is fitted with a two-way
stopcock for the introduction of the cathode mercury from the reservoir and also
for the withdrawal of the solution after the completion of the electrolysis. It is
closed with a Bakelite cover (fitted on to the top of the glass cell, which is ground
flat) and a gas delivery tube is provided for removing dissolved air from the
solution with nitrogen or other inert gas; the excess nitrogen escapes through the
loosely fitted glass sleeve through which the shaft of the glass stirrer passes.
Removal of the air is necessary, because oxygen is reduced at the mercury
cathode at about —0.05 volt vj. S.C.E., and this would interfere with the
deterimnation of most substances. The area of the mercury cathode is about
20 cm . Two kinds of anode, immersed directly in the test solution, may be used,
VIZ., a large helical silver wire {ca. 2.6 mm diameter, helix 5 cm long, and 3 cm
diameter; area about 100 cm^; as shown in the figure) or a platinum gauze

6 Volts

This may be improvised, as shown, from a portable galvanometer (for which the deflection is
directly proportional to the current) by inserting a large resistance in series with it.

545

XJll, 2 QUANTITATIVE INORGANIC ANALYSIS

cylinder (area 75 cm-) mourned vertically and co-axially with tlie stirrer shaft.
The silver anode is employed, inter alia, when the .solution contains metals, such
a.s bismuth, w'hich tend to be oxidised to insoluble higher o.xidcs at a platinum
anode; cliloride ion, at least equivalent to the quantity of the cathode reaction
(but preferably in 50-KK) per cent c.xcess) is added. The reaction at the silver
anode is

Ag-fCI =AgClTe

Hydra^'ine is med as the depolariser at the platinum iinode for metals which arc
not reduced by this compound:

NTl," ?irN: + 5ir +-le;

the evolution of nitrogen aids in removing dissolved air. The .salt bridge (4-msn
tube) from the saturated calomel electrode is iilled with 3 per cent agar gel
saturated with polassiiun chloride and its lip is placed within 1 mm of the
mercury cathode vvlieu tlie mercury is not being stirred; this ensures that the tip

Siifisii

bur

iraiLs in the mercury surface when tlic latter is stirred. It is essential that the
mercury-.soluiion interface (not merely the solution) be vigorously stirred, and
for this purpose the propeller blades of the glass stirrer are partially immersed in
the mercury.

The second type of mercury cathode cell, shown in Fig. Xlll. 4, utilises
magnetic stirring. A 250-cm^ Pyre.x beaker serves as the electrolysis vessel;
electrical contact with the cathode mercury is made by a short length of platinum
wire sealed into the side at the bottom or by means of a platinum wire sealed into
the bottom ot a glass tube and irnmer-sed in the mercury cathode. The stirring bar
floats on the mercury and results in smooth and cflicieni stirring of the
mercury-solution interlace. The anode is a stout platinum wire coiled into a flat

546

COULOMETRY Xin,2

spiral. The reference electrode is a silver-silver chloride electrode with its end just
barely brushing the surface of the mercury cathode; the top is held in a burette
clamp. It consists of a glass tube of about 10 mm internal diameter, the bottom of
which is closed by a sintered glass disc. The lower half of the tube is filled with a 3
per cent agar gel in saturated potassium chloride, and the upper half contains
saturated potassium chloride solution to which a drop of molar silver nitrate
solution has been added to saturate it with silver chloride. The electrode proper is
a length of pure silver wire, about 2 mm in diameter, dipping into the solution and
held in the rubber stopper. The potential of this electrode is +0.197 volt
(hydrogen electrode scale) .or -0.045 volt vs. S.C.E. When not in use the
electrode is stored with its lower end immersed in saturated potassium chloride
solution in a test tube. ,

One of the outstanding advantages of the mercury cathode is that the optimum
control potential for a given separation is easily determinable from polarograms
recorded with the dropping mercury electrode. This.potential corresponds to the
beginning of the polarographic diffusion current plateau; there is usually no
advantage in employing a control potential more than about 0. 1 5 volt greater
than the half-wave potential. Some values for the half-wave potential (E'lyj) and
suitable values for the cathode potential are collected in Table XIII, 1 . .

Table XIH, 1 Deposition of metals at controlled potential of the mercury cathode

Element

Supporting electrolyte

Volts vs. S.C.E.

Ei /2

^cathode

0.5M-acid sodium tartrate, pH 4.5

-0.09

-0.16

0.5Af-acid sodium tartrate, pH 4.5

-0.23

-0.40

0.5M-acid sodium tartrate, pH 4.5

-0.48

-0.56

IM-NH 4 CI+ IM-aq. NH 3

-0.81

-0.85

IM-NHiCl + lM-aq. NH 3

-1.33

-1.45

IM-pyridine + HCl, pH 7.0

-0.78

-0.95

1 A/-pyridine+ HCl, pH 7.0

-1.06

- 1.20

By means of the controlled cathode potential technique it is possible to effect
such difficult separations as Cu and Bi, Cd and Zn, and Ni.and Co. The
electrolysis is best conducted by using a potentiostat to automatically control the
potential of the mercury cathode at the desired value against a saturated calomel
or a silver-silver chloride reference electrode.

The general technique for performing a coulometric determination, at
controlled potential of the mercury cathode is as follows. -The supporting
electrolyte (50-60 cm^) is first placed in the cell and the air is removed by passing
a rapid stream of nitrogen through the solution for about 5 minutes. The cathode
mercury is then introduced through the stopcock at the bottom of the cell (Fig.
XIII, 3) by raising the mercury reservoir. The stirrer is started and the tip of the
bridge from the reference electrode is adjusted so that it just, touches, or trails
slightly in, the stirred mercury cathode. The potentiostat is adjusted to maintain
the desired control potential and the solution is electrolysed, with nitrogen
passing continuously, until the current decreases to a very small constant value
(the ‘background current’). This preliminary electrolysis removes ■ traces of
reducible impurities; the current usually decreases to 1 milliamp or less after
about 10 minutes. A known volume (say, 1 0-40 cm^) .of the sample solution is

547

XIII, 3/4 QUANTITATIVE INORGANIC ANALYSIS

then pipetted into the cell, anti the electrolysis is ailotvec! to proceed until the
current decreases to the same sntall value observed svith the supporting
electrolyte alone. Electrolysis is usually complete within an hour. The
hydrogen-oxygen coulometer is then read, and the weight H'of metal deposited
is calculated from the e.xprc.ssion

ft'

iiF

wlicre .1/ is tlic atomic weight of the metal. Q i,s the total quantity of electricity
(coulombs) deduced from the reading of the coulometer (or current-time
integrator), /j is the linal background current (amperes), i the electrolysis time
(seconds), is the number of electrons required for the reduction, and F is the
faraday constant. In many cases the correction for llie background current is
negligible and the factor /^r may be neglected.

Experimental details (in outline) for the sepani tion of nickel and cobalt follow.

xni, 3. SKPAU.VnON OF NICKEL AND COBALT BY COULO.METUIC
ANALYSIS AT CON TROLLED POTEN I f Al.. Reauents, Suuuiard nickel-
and cohah-ion .suhitiun.s. Prepare .standard solutions of nickel and cobalt ion
(cu. lOmgperem-’) from pure ammonium nickel sulpliatc and pure ammonium
cobalt sulphate respectively.

Pyridine. Redistil A.R. pyridine and collect the middle fraction boiling within
a 2’ range.

Suppariiny eleciralyte. Prepare a supporting electrolyte composed of l.OO.U*
pyridine and 0.50,)/-chloridc ion. adjusted to a pH of 7.0±0.2 for use with a
silver anode, or I.003/-pyridine. O.jO.If-chloride ion and 0.2{).I/*hydrazinium
sulphate, adjusted to a pll of 7.0 ±0.2, for li ve with a platinum c.ithode. A small
background current is obtained with the latter.

Procedure. Place 90 cm' of the supporting electrolyte in iheceil (Fig. Xill,
.i), remove dissolved air with pure nitrogen, and subject the solution to a
preliminary clecirolysi.s with the potential of the mercury cathode - 1.20 volt rj.
S.C.E. to remove irace.s of reducible impurities; stop the electrolysis when the
background current Ua. 2 milliamp) has decreased to a con.stanl value (30-60
minutes). Prepare the coulometer. adjust the polentiostat to maintain the
potential of the cathode at the value to be used in the determination ( — 0.95 volt
w. S.C.E. for nickel and - 1.20volt vxS.C.E. for cobalt), and add 20.0 cm' of the
sample solution. Continue the electrolysis under automatic control until the
current decreases to a constant minimal value (2~3 hours foreach metal). Record
the total quantity of cleciriciiy passed (evaluated from the coulometer readings),
the electrolysis time, and the final current.

Calculate the weight of metal deposited at each potential using the relation
given in Section Xlll, 2.

Coulometry at constant current: coulometric titrations

XIII, A. GENERAL DISCUSSION. Coulometry at controlled potential is
applicable only to the limited number of substances which undergo quantitative
reaction al an electrode during electrolysis. By using coulometry at controlled or
constant current, the range of substances that can be determined may be
e.xtended considerably, and includes many which do not react quantitatively al

548

COULOMETRY XIII, 4

an electrode. Constant-current electrolysis is employed to generate a reagent
which reacts stoichiometrically with the substance to be determined. The
quantity of substance reacted is calculated with the aid of Faraday’s law, and the
quantity of electricity passed can be evaluated simply by timing the electrolysis at
constant current. Since the current can be varied from, say, 0. 1 to 100 milliamp,
amounts of material corresponding to 1 x 10“®to 1 x 10“® equivalent per second
of electrolysis time can be determined. In titrimetric analysis the reagent is added
from a burette; in coulometric titrations the reagent is generated electrically and
its amount is evaluated from a knowledge of the current and the generating time.
The electron becomes the standard reagent. In many respects, e.g., detection of
end-points, the procedure differs only slightly from ordinary titrations.

The fundamental requirements of a coulometric titration are: (i) that the
reagent-generating electrode reaction proceeds with 100 per cent efficiency, and
(ii) that the generated reagent reacts stoichiometrically and, preferably rapidly,
with the substance being determined. The reagent may be generated directly
within the test solution or, less frequently, it may be generated in an external
solution which is allowed to run continuously into the test solution.

Since a small quantity of electricity can be readily measured with a high degree
of aecuracy, the method has high sensitivity. Coulometric titrimetry has seyeral
important advantages :

1. Standard solutions are not required and in their place the coulomb becomes
the primary standard.

2. Unstable reagents, such as bromine, chlorine, silver(II) ion (Ag^"*"), and
titanium(III) ion, can be utilised, since they are generated and consumed
immediately ; there is no loss on storage or change in titre.

3. When necessary very small amounts of titrants may be generated: this
dispenses with the difficulties involved in the standardisation and storage of
dilute solutions. The procedure is ideally adapted for use on a micro- or
semimicro-scale (e.g.. Ref. 8).

4. The sample solution is not diluted in the internal generation procedure. ' -

5. By pre-titration of the generating solution before the addition of the sample,
more accurate results can be obtained. The end-point indicator corrections
are thus automatically cancelled and the effect of impurities in the
generating solution is minimised.

6. The method (which is largely electrical in nature) is readily adapted to remote
control: this is significant in the titration of radioactive or dangerous
materials. It may also be adapted to automatic control because of the relative
ease of the automatic control of current.

Seyeral methods are available for the detection of end points in coulometric
titrations. These are :

(a) Use of chemical indicators: these must not be electro-active. Examples
include methyl orange for bromine, starch for iodine, dichlorofluorescein for
chloride, and eosin for bromide and iodide.

(b) By potentiometric observations. Electrolytic generation is continued until
the e.m.f. of a reference electrode-indicating electrode assembly placed in the
test solution attains a pre-determined value corresponding to the equivalence
point.

These are based upon the establishment of conditions such that either the
substance bemg determined or, more usually, the titrant undergoes reaction at an

549

Xm,4 QUANTITATIVE INORGANIC ANALYSIS

indicator electrode to produce a current whicit is proportional to the
concentration of the electro-active substance. With the potential of the indicator
electrode maintained constant, or nearly so. the end point can be established
from the course ol the current citan^c duriitg the titr.ition, T he voltaj^e impressed
upon the indicator electrode is well below the 'decomposition voltage’ of the pure
supporting electrolyte but close to or abo\e the 'decomposition voltage’ of the
supporting electrolyte plus Ircc titraiit, consequently, as long as any of the
substance being determined remains to react with the tiuant, the indicator
current icmains very small but increases as soon as tltc end point is passed and

free tiirant is present. There is a relatiscly incshaiislible supply of titrant ion Ic.g,

bromide ion in eoulometric titrations with bromine), and the indicator current
beyond the equivalence point is therefore governed largely by the rate of
dilfusion of tiie free litrant (c.g.. bromine) to llie surface of the indicator
electrode. The indicator current is eonscqueniiy proportional to the
concentratiun of the free tiirani (eg-, bromine) in the bulk of the solution and to
the area i)f tlie indicator electrode (cathode lor bromine). The indicator current
will increase with ineieasing r.iie of stirring, .since this decreases the thic-kncss of
the dilTu-sion layer at ilie elcetoHle; it is also ssunewhat tcinperature-depciidenl.
The generalion time at winch the equivalence point is reached may Ixidctentiined
by calibrating the mdic.itor clcciriidc system with the .siipponing cleciroijic
alone by generating the mr.int (c.g., bromine) ('or various times (say, 10-50
seconds) to evaluate the constant iti the relation /, Ki. whete /, is the indicator
current and / is tlie time. The generating time to the equivalence point may (hen
be obtained from the otwerved linal v.iiiie ofihe indicator current inlheactiu!
titration, calculating the evecss itcuer.iting time aiui subtracting this from ihi
total generating time in liie titration. Alternatively, and more simply, ihr
equivalence point time may be located by me.i-suring three values of the indicator
current at three me.isured times beyond the equivalence point and cvtrapolaling
to zero curreiu-

(</) By application of the bi.imperumetric (dead-stop) method (comp-uc Section
XVH, 14),

The titration cell consists of .i specirophois'iiieter cuvette (2cin light path).
The motor-driven glass piopeiler siirier and the working platinum elaircdcare
placed in tliecell in sueii .i w.iy as to beout of the light pallita piatinumclcctrodi
in dilute sulphuric acid in an adjacent cineiiealso placed in the cell holder .senes
us an auxiliary electrode and is connected w ith the titr.ition cell by .in inverted U-
tube .salt bridge. The .ippropn.ite wavelength is set cm the insiruracnl. Iklorethe
end point the absorbance changes only very slowly, hut a (.ipid and linear
response cveurs Iseyond the equivalence point. Mvample.s are; the titration ol
I’e(ll) in dihiic .sulphuric acid with electro-generated Cc(lV) at 400 nni. and the
titration ot arsenic! I II) with elect ro-gcncratcsl iodine at .v)2niii.

I he principle of eoulometric liiratioii, involving the generation ofa titr.iniby
electrolysis, may be illustrated by reference tsi the titration of itondli wth
clcclro-gencralcd ccriiim(lV). .A l.irge excess of Cc(lll) is added to (he solution
containing the Fc(ll) ion m the presence of. stiy LU-su!phuric acid. Let us
consider what hapjieiis at .i platinum anode when a solitlion containing Fc-(ll)
ions alone is electrolvscci at constant current. Initially the waction
be* ^ ci Pc' ' -r <’ will proceed with 100 per cent curtciu eliiciency, ,Al the anode
surface the concentration of l'e(UI) ions formed is relatively large, while that of

.“iSO

COULOMETRY XIII, 5

the Fe(II) ions, which is governed -by the rate of transfer from the bulk of the
solution, is very. small: the potential of the anode gradually acquires, a value
which is much more positive (more oxidising) than the standard potential of the
Fe(III)/Fe(II) couple (0.77 volt). As electrolysis proceeds, the anode potential
becomes more and more positive (oxidising) at a rate that depends on the current
density, and ultimately it becomes so positive (ca. 1.23 volt) that oxygen
evolution from the oxidation of water begins ( 2 H 2 O = O 2 +4H''' +4e), and this
occurs before all the Fe(II) ions in the bulk of the solution is oxidised. As soon as
oxygen evolution commences, the current efficiency for the oxidation of Fe(II)
falls below 100 per cent and the quantity of Fefll) initially present cannot be
computed from Faraday’s law. If the electrolysis is conducted in the presence of a
relatively large concentration of Ce(III) ions the following reactions will take
place at the anode. At a certain potential of the anode, which is considerably less
than that required for oxygen evolution, oxidation of Ce(III) to Ce(IV) sets in,
and the Ce(IV) thus produced is transferred to the bulk of the solution, where it
oxidises Fe(II). The potential of the working electrode is thus stabilised by the
reagent-generating reaction, and hence is prevented from drifting to a value such
that an interfering reaction may result. The resulting Ce(IV) ions readily react
with the remaining Fe(II) ions, according to the reaction :*

Ce‘^+-t-Fe^+ =Ce3"'-l-Fe^ +

Stoichiometrically, the total quantity of electricity passed is exactly the sarne as it
would have been if the Fe(II) ions had been directly oxidised at the anode and the
oxidation of Fe(II) proceeds with 1 00 per cent efficiency. The equivalence point is
marked by the first persistence of excess Ce(IV) in the solution, and may be
detected by any of the methods described above.

Side reactions are avoided at the generating electrode provided there is not
complete depletion (at the electrode surface) of the substance involved in the
generation of the titrant. The concentration of the titrant depends upon the
current through the cell, the area of the generating electrode, and the rate of
stirring; the concentration of the generating substance is usually between O.OIM
and 0.1 M.

Xin, 5 . INSTRUMENTATION. Constant current sources. The currents
used in coulometric titrations are usually in the range 1-50 milliamp. Fairly
constant currents are conveniently obtained from batteries with a series
regulating resistance ; seven 6-voltcar (or storage) batteries in series yielding the
equivalent of a 42-volt battery will be found satisfactory. Periodic adjustment of
the series resistance may be required to maintain constant current..

* The oxidation of Ce(III) ions at the surface of the electrode probably proceeds, at least in part by
the reaction;

Ce^ -^ + 380*^ - [Ce(SO.,)3]" - e

the cerium complex thus produced reacts with Fe(II) ions in the solution:

[Ce(S04)3]2- + Fe^+ = Ce^+ + 3S04^--t-Fe^'’'.

The net reaction is

■Fe^+^Fe^+-(-e. , . . ■

Regardless of the actual path of the reaction, one mole of Fe(II) is oxidised by each faraday
Constant. '

551

Xm, 6 QUANTITATIVE INORGANIC ANALYSIS

If Vg is the voltage provided by the batteries {ca. 42 volts), V(. is the voltage
across the cel) (1--2 volts), R the scries regulating resistance (maximum value, say,
10000 ohms), Rt, the internal resistance of the battery itself, and R^- the resistance
of the cell (probably not greater than about 20 ohms), the current I (lowing
through the circuit is given by:

r+'Ri^+r;

The maximum variation of the cel! voltage is of the order of 0.5-0.7 volt. For
Vg = 42 volts, the maximum variation is about 1 per cent. To maintain the
batteries in good condition and the .series rcstsiance in thermal equilibrium, it is
advisable to employ a switching arrangement whereby the titration cell is
replaced by a dummy resistance during the intervals between titrations. The
dummy resistance is .selected so that tliecurrem will have about thesainc value as
during titrations.

More rigorous control of the current may be achieved by the use of
commercially available stabilised power unit.s, but better, by the use ofa purpose-
designed constant current source (e.g.. as in Ref. 4). or a commercial 'amperostaf
(Section XIH,2).

Current-measuring detices. The obvious means for measuring the current
IS a carefully calibrated milliammcter; a more precise method is to detennine
with a good pytenttonieter the voltage drop /f, acrt)s.s a precision rc,sistance
(50-100 ohms, A’l ) in series with the electrolysis cell. The electrolysis current /,. is
calculated from the equation/,

Time nicasureuieut. An ordinary electric stop-dock operated by closing
and opening the electrolysis circuit is not very satisfactory because of the
appreciable lag and ‘coa.st’ of the motor: the error may amount to several tenths
of a sa'ond. The beat type of electric stop-cliK-k is one fitted with magnetic
brakes; this starts and stops simultaneou.sly with the starting and stopping of the
current. Sucii electric .stoi>ciock.s (c.apacity lOOl) seconds or more) are available
commercially, and readings may be made precise to 0.01 second per operation.
Short-period variaiioirs in the frequency of the a.c. mains supply can cause
significant errors in the measurement of lime intervals of several minutes
duration ; the error may be eliminated by the use ofa frequency-regulated power
supply for the clock. The electric timer should be controlled by the same switch
which starts and stops the electrolysis current.

A method which is convenient, but which does not yield results of the highest
precision, is to include an integrating electric motor in the circuit, and driven by
the voltage drop across a precision fi.xed resistance in scries with the electrolysis
cell; it is fitted with a counter which gives the product, current x lime. An
integrating milliammcter may also be employed ( Ref. vS).

For the highest precision, a quartz crystal clock may be used (Ref. 4), or a
dekatron timer ( Ref. 5).

XIII, 6. CIRCUIT AND CELL FOR COULOMETRY .A T CONTROLLED
CURREN r. Fig, XllI, 5 is a schematic diagram of the circuit for coulomctric
titration with internal generation of the titranl u.sing the dead stop or
amperometric end-point technique. The d.c. supply may be obtained from a
bank of storage batteries (accumulators) delivering 42 volts; the two variable

552

COULOMETRY Xin,6

high-wattage resistances (large) and R 2 (small) permit the current to be
varied. Alternatively, and more conveniently, an electronically controlled
current supply unit may be used (see Section XHI, 5). The calibrated
milliammeter M records the generating current; a more accurate value of the
current is obtained by measuring the voltage drop across a standard resistance R^
(say, 100 ohms) with a potentiometer P. The variable resistance (high-wattage
type) is so connected that the electrolysis current flows through it whenever the
electrolysis cell is disconnected from the circuit; its value {ca. 20 ohms) should
not differ greatly from that of the cell, and is easily adjusted by arranging that the
current will have nearly the same magnitude as in the titration. This arrangement
ensures that the resistances and R 2 are at constant temperature, and so
minimises the variations in their resistance which would occur if the current
through them were interrupted periodically.

The electrolysis cell contains the working or generator electrode A, at which
the reagent is electro-generated, and the auxiliary electrode C. Electrode A may
be of platinum, silver, or mercury (the last-named will, of course, be a pool at the
bottom of the cell) ; electrode C is usually of platinum. The auxiliary electrode C
is generally placed in a separate glass tube closed at its lower end by a fine-
porosity glass disc: the level of the solution in this compartment must be
maintained at a higher level and greater ionic strength than the solution in the
titration cell, so as to prevent diffusion of the latter into the isolated
compartment. The electrolyte in which C is immersed may be either the same as
the supporting electrolyte in the test solution or some other innocuous electrolyte
appropriate to the particular case.

. and E 2 are the indicator electrodes. These may consist of a tungsten pair for
a biamperometric end-point; for an amperometric end-point the indicator
electrodes may both be of platinum foil or one can be platinum and the other a
saturated calomel reference electrode. The voltage impressed upon the indicator
electrodes is supplied by battery B (ca. 1 .5 volts) via a variable resistance R ^ ; N
records the indicator current. For a potentiometric end-point and E 2 may
consist of either platinum-tungsten bimetallic electrodes, or E^ may be an S.C.E.
and £2 a glass electrode. These are connected directly to a pH meter with a
subsidiary scale calibrated in millivolts.

553

COULOMETRY XHI, 6

Indicator

high-wattage resistances (large) and i ?2 (small) permit the current to be
varied. Alternatively, and more conveniently, an electronically controlled
current supply unit may be used (see Section XHI, 5). The calibrated
milliammeter M records the generating current ; a more accurate value of the
current is obtained by measuring the voltage drop across a standard resistance i ?3
(say, 100 ohms) with a potentiometer P. The variable resistance R^ (high-wattage
type) is so connected that the electrolysis current flows through it whenever the
electrolysis cell is disconnected from the circuit ; its value {ca. 20 ohms) should
not differ greatly from that of the cell, and is easily adjusted by arranging that the
current will have nearly the same magnitude as in the titration. This arrangement
ensures that the resistances R^ and R 2 are at constant temperature, and so
minimises the variations in their resistance which would occur if the current
through them were interrupted periodically.

El and E 2 are the indicator electrodes. These may consist of a tungsten pair for
a biamperometric end-point; for an amperometric end-point the indicator
electrodes may both be of platinum foil or one can be platinum and the other a
saturated calomel reference electrode. The voltage impressed upon the indicator
electrodes is supplied by battery B (ca. 1.5 volts) via a variable resistance R ^ ; N
records the indicator current. For a potentiometric end-point E^ and £3 t^^y
consist of either platinum-tungsten bimetallic electrodes, or may be an S.C.E.
and £2 a glass electrode. These are connected directly to a pH meter with a
subsidiary scale caUbrated in millivolts.

553

Xnr, 6 QUANTITATIVE INORGANIC ANALYSIS

S is a cloublo-poic toggle switch. 'Fhis permits the simultaneous operation of
the electrolysis current and the electric timer T; when the current is not passing
through the cell, it passes through the equivalent resistance R^.

A more detailed drawing of the titration cell is shown in Fig, Xfll, 6. It consists
of a tail-form beaker (without lip) of about 1 50 or 200 cm^ capacity. Provision is
made for magnetic stirring and for passing a stream of inert gas (e.g., nitrogen)
through the solution. The main generator electrode A may consist of platinum
foil ( 1 X 1 cm or 4 X 2,5 cm) and the auxiliary electrode C may consist of platinum
foil (1x1 cm or 4 x 2.5 cm) bent into a half cylinder so as to fit into a wide glass
tube (cn. 1 cm diameter). The isolation of the auxiliary generator electrode C
within the glass cylinder (closed by a sintercd-glass disc) from the bulk of the
solution avoids any elTects arising from undesirable reactions at this electrode.
The nature of the indicator electrodes £, and E, will depend upon the procedure
adopted for the detection of the end-point— -biamperometric, amperometric. or
potentiomctric -as described above. In general, the indicator electrodes should
be positioned outside the electric field (current path) between the generator
electrodes, otherwise spurious indicator currents may be produced, particularly
in the amperometric detection of the equivalence point.

sw >

if an integrating motor is used, the
circuit shown in Fig. XT 11, 5 is modified
as shown in Fig XHl. 7, The right-hand
side of the diagram is similar to that of
Fig. .Xlll, 5. with the obvious omission
of the timing device; the milliammcter
need not be calibrated. The integrating
motor-counter unit IM is connected
across the high stability resistance SR.
The electrolysis cell contains two gener-
ating electrodes and two indicator
electrodes; the latter arc connected to a

Fi ' XIII 7 device (denoted by liulkutor) for the

’ detection of the equivalence point by

biamperometric, amperometric, or potentiometrie methods. SW is a switch,
usually of the double-pole type.

Before use, the integrating motor must be calibrated by passing an accurately
measured constant current through a high-stability resistance for an accurately
measured time and reading the corresponding count on the integrating motor.
The calibration factor may be expressed as:

= Coulombs percount ^

Number ol counts

The general procedure for making a determination is as follows: The
electrolysis cell is set up with both generator and indicator electrodes in position
and provision is made, if necessary, for passing an inert gas (c.g„ nitrogen)
through the solution. The titration cell is charged with the solution from which
the titrant will be generated elcctrolytically. together with the solution to be
titrated. The auxiliary electrode compartment is filled with a solution of the
appropriate eieclrolyie at a higher level than the solution in tiie titration cell. The
indicator electrodes are connected to a suitable apparatus for the detection of the
end-point, e.g,, a pH meter with additional millivolt scale, or a galvanometer.

554

COULOMETRY Xin,7

Stirring is effected with a magnetic stirrer. The reading of the digital indicator
instrument is taken. The current, previously adjusted to a suitable value, is then
switched on and reaction between the internally generated titrant and the test
solution allowed to proceed. Readings are taken periodically (more frequently as
the end point is approached) of the integrating motor counter and of the
indicating instrument (e.g., pH meter); it is usually necessary to switch off the
electrolysing current while the readings of the indicating instrument are
recorded. The end-point of the titration is readily evaluated from the plot of the
reading of the indicating instrument (e.g., millivolts) against the counter readiiig ;
the first or second derivative curve is drawn to locate the equivalen9e point
accurately. It is possible to repeat the titration with a fresh volume of the test
solution. If the end point is determined potentiometrically subsequent
determinations may be stopped at the potential found for the equivalence point
in the initial titration.

Xffl, 7. EXTERNAL GENERATION OF TITRANT. The limitations of
coulometric titration with internal generation of the titrant include :

1. No substance may be present which undergoes reaction at the generator
electrodes ; for example, in acidimetric titrations the test solutions must not
contain substances which are reduced at the generator cathode.

2. When applied on a macro scale — samples of 1-5 milli-equivalents —
generation rates of 100-500-milliamp are required: parasitic currents are
induced in the indicator electrodes at currents in excess of about 10-20
milliamp, consequently precise location of the equivalence point by
amperometric methods is not trustworthy.

To overcome these limitations, the reagent can be generated at constant
current with 100 per cent efficiency in an external generator cell and subsequently
delivered to the titration cell. This technique is identical with an ordinary
titration except that the reagent is generated electrolytically. A double-arm
electrolytic cell for external generation of the titrant is shown in Fig. XIII, 8. The
generator electrodes consist of two small platinum spirals near the centre of the
inverted U-tube. The space between the electrodes is packed with glass wool to
prevent turbulent mixing; the downward legs of the generator tube are
constructed of 1-mm capillary tubing to reduce the inconvenience due to. hold-
up. The solution of the electrolyte, which upon electrolysis will yield the desired
titrant, is fed continuously into the top of the generator cell. The solution is
divided at the T joint so that about equal quantities flow through each of the arms
of the cell. As these portions of the solution flow past the electrodes, electrolysis
occurs : the products of electrolysis are swept along by the flow of the solution
through the arms and emerge from the delivery tips. A beaker containing the
substance to be titrated is placed beneath the appropriate delivery tip, arid the
solution from the other tip is run to waste. Thus for the titration of acids,
electrode a functions as a cathode in sodium sulphate generator electrolyte and
the hydroxide ion generated by the reaction 2H20-t-2e = 20H“ +H2 flows into
the test solution. The hydrogen ion and oxygen generated at the other electrode
by the reaction 2H2O = 4H'^ + O2 -1- 4e are swept out of the other arm into the
drain. For the titration of bases, the generator electrode which delivers to the
titration cell is employed as the anode. For titrations with electrically generated
iodine, the generator electrolyte consists of potassium iodide solution, and the
iodine solution formed at the anode flows into the titration vessel.

555

XIII, 8 QUANTITATIVE INORGANIC ANALYSIS

From gciKTator
clcclroljlc
supply tcscr\oir

A minor ciisaclvaniaizc of external generation of titrant is the dilution of the
contents of tlie titration cell ; care is therefore necessary in suitably adjusting the
rale of How and the concentration of the generator solution. The procedure is,
however, admirably suited for automatic control.

Xni, 8. EXPERIMENTAL DETAILS FOR TYPICAL COULOMETRIC
TITRATIONS AT CONSTANT CURRENT. In the following pages
experimental details will be given for some typical coulometric titrations at
constant current. Either of two procedures for determining the total quantity of
electricity passed may be used:

1. Maintenance ofcoastant current and exact measurement ofthe time during
which current i.s passed. The current may be read on a calibrated milliammeter or
can be evaluatetl with the aid of a potentiometer and standard resistance. The
time can be measured with au electric timer provided with solenoid brakes and
operated from the a.c. supply mains. Results of nioderatc accuracy can usually be
obtained by the use of ;i good stop-clock or stop-watch.

2. Maintenance of a reasonably constant current coupled with the use of a
low-inerlia integrating motor; this technique is generally more convenient but is
somewhat le.ss accurate.

Alternatively, a commercial coulometric titraior may be employed;'' for
details of precision measurements. Refs. (l)-{6) should be consulted. It is
considered that, with appropriate attention to detail, coulometric analysis is
probably capable of better precision than any other technique, and the
suggestion has been made that the Farad;iy constant should be regarded as the
prime chemical standard.

By virtue of its inherent accuracy, coulometric titration is very suitable tor the

* Suppliers include, inter alia, l>oliriii;iiiii-En\irotcch: XtcKec I’edeiscii In.suunicnts: Mctrohni,
Hcrisau; Princeton Allied Research; Radiometer Ltd.

556

COULOMETRY Xni,9

determination of substances present in small amount, and quantities of the order
of 10“’ to 10"^ of a mole are typical. Larger amounts of material require .very
long electrolysis times unless an amperostat capable of delivering relatively large
currents (up to 2 A) is available. In such cases, a common procedure is to start the
electrolysis with the heavy duty apparatus, and then to switch to one with a much
lower output as the end point is approached.

In the determinations described below, use of an integrating motor is assumed ;
the modifications necessary if other equipment is used should be evident.

Xm,9. ANTIMONY(III). Discussion. Iodine (or tri-iodide ion

I3 - ^ I2 4- 1“) is readily generated with 100 per cent efficiency by the oxidation of
iodide ion at a platinum anode, and can be used for the coulometric titration of
antimony(III). The optimum pH is between 7.5 and 8.5, and a complexing agent
(e.g., tartrate ion) must be present to prevent hydrolysis and precipitation of the
antimony. In solutions more alkaline than pH of about 8.5, disproportionation
of iodine to iodide and hypoiodite (or iodate) occurs. The reversible character of
the iodine-iodide complex renders equivalence-point detection easy by both
potentiometric and amperometric techniques; for macro titrations, the usual
visual detection of the end point with starch is possible.

Apparatus. Use the apparatus shown in Figs. XIII, 5, 6 and 7. The
generator cathode (isolated auxiliary electrode) consists of platinum foil
(4x2.5 cm, bent into a half cylinder) and the generator anode (working
electrode) is a rectangular platinum foil (4x2.5 cm). For potentiometric end-
point detection, use a platinum-foil electrode 1.25 x 1.25 cm (or a silver-rod
electrode) in combination with a saturated calomel reference electrode connected
to the cell by a potassium chloride- or potassium nitrate-agar bridge.

Reagents. Supporting electrolyte. Prepare a O.lM-phosphate buffer of
pH = 8 containing O.lM-potassium iodide and 0.025M-potassium tartrate
(say, 0.17g of Na2HP04,12H20, 3.38 g NaH2P0.^,2H20, 4.15 g KI and 1.5 g
potassium tartrate, all A.R. salts, in 250 cm^ of water).

Antimony potassium tartrate, 0.0 IM.; use the dried A.R. salt.

Procedure. Place 45 cm^ of the supporting electrolyte in the cell and fill the
isolated cathode compartment with the same solution to a level well above that iii
the cell. Pipette 5.00, 10.00, or 1 5.00 cm^ of the 0.0 IM-antimony solution into the
cell and titrate coulometrically with a current of 40 milliamp. Stir the solution
continuously by means of the magnetic stirrer and take e.m.f. readings of the
Pt-S.C.E. electrode combination at suitable time intervals: the readings may be
somewhat erratic initially, but become steady and reproducible after about three
minutes. Evaluate the end point of the titration from the graph of e.m.f. vj.
counter reading; this will be similar in shape to the curve shown in Fig. XIII, 9
(Section XIII, 12), but in the reverse order. If it proves difficult to locate the end-
point precisely, recourse may be made to the first and second differential plots.

If it is desired to use the biamperometric method for detecting the end-point;
then the calomel electrode and also the silver rod (if used) must be removed and
replaced by two platinum plates 1.25 cm x 1.25 cm. The potentiometer (or pH
meter) used to measure the e.m.f. must also be removed, and one of the indicator
dectrodes is then joined to a sensitive galvanometer fitted with a variable shunt.
The indicator circuit is completed through a potential divider placed across a
1.5 V dry battery (see Fig. XVII, 5 ; Section XVH, 14). Charge the electrolysis cell
as described above, adjust the potential across the indicator electrodes to about

557

Xin. lO/l I QUANTITATIVE INORGANIC ANALYSIS

150mV, and set the galvanometer shunt to give maxiinum deflection on the
galvanometer. Switch on the electrolysis current and read the indicator current
from time to time. Plot indicator current against counter reading and c.xtrapolate
to zero current to locate the end point.

For determination of the end-point by a visual method, add l-2cm^ of 1 per
cent .starch .solution, and stop the titration immediately the solution has acquired
a uniform blue colour.

XIII, 10. THIOSULPHATE. Dheussion. Thiosulphate may be titrated
coulometricaily with electro-generated iodine, using starch for visual end-point
detection. As the end point is approached, deep blue streaks appear which
spread into the solution upon stirring: at the end point the anolyic suddenly
acquires a uniform blue colour. Care must be taken that the titration is stopped at
the very first darkening in colour of the test solution, otherwise high results are
obtained.

Apparatus. Sec Section Xlll, 9.

Reagents. Xuppor/h/g electrolyte. Dissolve 0.5 g pure potassium iodide in
40 cm^ water.

Sodium ihioxulpluite, 0.0/M. Prepare from the .A.R. salt using boiled-oul
water.

Catholyte. A 1-2 per cent potassium chloride solution acidified faintly with
dilute hydroehloric acid.

Procedure. Place 40cni^ of the supporting electrolyte in the cell, together
with 5.00, 10.00, or IS.OOcm^ oftheO.Ol.U-thiosulphaie solution and l-2cm^of
starch solution. Fill the isolated cathode compurlinenl with the acidified
potassium chloride solution. Stir the solution magnetically. Pass a current of 30
milliamp until the anolyte first acquires a uniform dark colour.

Xlll, 11. OXINE (8-HYDROXYQULNOLlNE). Dhcu.x.sion. Bromine may
be electro-generated with 100 per cent current cfiiciency by the o.sidalion of
bromide ion at a platinum anode. Broinination of o.xinc proceeds according to
the equation;

C.,H 70 N + 2Br;= C,,HjONBr,!-2ir +2Br-

and thus four faraday constants arc required per mole of o.xine. The end-point is
detected amperometrically,

Apparatas. Set up the apparatus as in Section XIII, 9 with two small
platinum plates connected to apparatus for the amperomctric detection of the
cnd-poinl.

Reagents. Supporting electrolyte. Prepare 0.2.t/-potassium bromide
from the A.R. salt.

O.xiiie .wltition. 0.003, \/-o.\inc (use the A.R. material) in 0.00253/'-
hydrochloric acid.

Procedure. Place 40em’’ of the supporting ciecirolyte in the coulometric
cell and pipette lO.OOein^ of the o.xinc .solution into it. Charge the cathode
compartment with the 0.2.'\/-potassium bromide. Pass a current of 30 milliamp
while stirring the .solution magnetically. Adjust the sensitivity of the indicating
apparatus to a suitable value. Near the end point transient deflections occur and

558

COULOMETRY XIII, 12

serve to give warning of its approach. The end point is at the first permanent
deflection and the reading of the counter is taken. '

Xni, 12. POTASSIUM DICHROMATE (DICHROMATE ION)

Discussion. Iron(II) ions are generated electrolytically by reduction of
iron(III) ions at a smooth platinum cathode. They reduce dichromate ions
present in the same solution; the equivalence point is determined
potentiometrically using a platinum-tungsten electrode pair.

Fe^'*' +e^Fe^’^

,6Fe^+ + Cr 207 =^- + 14H+ = dFe^^ +2Cr3+ + 7 H 2 O

Apparatus. Set iip the circuit and the electrolytic cell (as in Section XIII, 9) .
Fill the isolated anode compartment with ca. 0.2A/-sodium sulphate (65 g of
A.R. Na2S04,10H20 per litre) and maintain its level above that of the solution in
the cell. Use a platinum sheet (1.2cm^) and a tungsten helix as indicating
electrodes. Measure the potential change by means of a digital voltmeter or a pH
meter provided with an additional millivolt scale.

Reagents. Ammonium iron (III) sulphate solution, ca. 0.5M. Dissolve
145g A.R. ammonium iron(III) sulphate in 125 cm^ water containing lOcm^
concentrated sulphuric acid in a 600-cm^ beaker, add 45 cm^ of concentrated
sulphuric acid cautiously, followed by 15cm^ 100-volume hydrogen peroxide.
Maintain the solution at 50-70 °C until evolution of oxygen ceases {ca. 30
minutes). When cold, filter through a fine-porosity sintered-glass funnel and di-
lute to 500 cm^. The above treatment removes any iron(II) which may be present.

Sulphuric acid, ca. 9M.

Potassium dichromate solution, O.IN. Prepare a O.IW solution from the dried
powdered A.R. salt. Prepare also a more dilute solution, e.g., 0.01 N, by dilution.

Procedure. Pipette the dichromate solution (say, 5.00 cm^) into the
titration cell, add 2 cm^ 9M-sulphuric acid, and then the ammonium iron(III)
sulphate solution (25 cm^ for O.liV- or 10 cm^ for 0.01N-K2Cr2O7). Dilute the
solution in the cell sufficiently to cover the electrodes. Pass pure nitrogen through
the solution for about 15 minutes in order to remove dissolved oxygen and
continue the passage of gas during the eleetrolysis.

Adjust the current before the titration to the desired value (15-20 or 40-50

Counter readings

Fig.Xin,9

milliamp). Commence the electrolysis and
record the counter reading. Follow the
potential change across the indicator
electrodes on a digital voltmeter or on a pH
meter (millivolt scale) as the titration
proceeds. Interrupt the electrolysis current
and the gas stream momentarily while
taking voltmeter readings; this procedure
is essential near the end-point to allow the
system to reach equilibrium. The voltage
change {ca. 200 millivolts) is abrupt at the
stoichiometric end point. The latter can be
evaluated precisely by plotting a voltage-
count curve and the first or second differen-
tial derived therefrom. A typical titration
curve (for 5.00 cm^ of 0.01.iV-K2Cr207)

559

xni, 13/14 QUANTITATIVE INORGANIC ANALYSIS

is shown in Fig. XIII, 9. Having once located the end-point potential, subsequent
titrations may be stopped at this value; it is essential, however, to de-o.xygcnate
the solution before each titration and to maintain the level of the solution in the
isolated anode compartment above that in the cell.

XIll, 13. IRONfll). Discussion. Cerium(lV) ions are generated at a bright
platinum anode from a supporting electrolyte containing a high concentration of
sulphuric acid and of Cedll): the equivalence point is determined potentio-
metrically. The oxidation potential of the Ce(lll) -Ce(IV) couple in sulphuric
acid is 4- 1 .43 volt and is relatively close to the potential at which water is o.xidised
at a platinum anode (2H, 0x^0,+ 411 ’’ +4<.’). The Ce(lV) is generated in
a solution containing a high concentration of Cc(III), and consequently
the generator anode operates at a potential well below the standard potential.

Apparalas. Uj>e the apparatus de.scribcd in Section XIII, 9. except that the
generating electrodes are reversed, i.e., the auxiliary cathode is in the fsolated
cathode comparlmeni containing 15 per cent ammonium sulphate solution or
I.5.lf-sulphuric acid. The end-point is determined poieiuiomeirically using a
platinum indicator electrode and a saturated calomel reference elccirodc.
RcagonLs. Sulphuric uciJ. 9M aiul /,5M.

Ceriwnl III) .\u!phah' soluiioii. ea. d./.\l. Dissolve cerium(lll) sulphate or,
better, ammonium ceriumdil) sulphate Clow in rare earths') in boiled-out
distilled water.

Iroit(Il) solution, O.OIM. Prepare a 0. 14f ammonium ironlM) sulphate
solution from the A.R. .salt, a little .V/-sulphurie acid, and boiled-out distilled
water. Dilute svith boiled-out distilled water to 0.01.1/ concentration.

Proccdurt:. Place 40cnd of the ceriumdil) solution and lOcm^ of 9.1/-
sulphurie acid in the titration cell, and 1.5.l/-sulphurie acid in the isolated
catliode comparlmeni (the Ie\el mu.si be above that of the ultimate level in the
main cell). Pass nitrogen for 10 minutes to remove dis.sohed air and maintain the
stream of gas during the titration. Pipette S.OOcnd of the ironfilj solution into
the coulomctric cell, and adjust the level ofihe liquid in the cathode compartment
by adding l.5.1/-.sulphuric acid with a dropper pipette. Titrate coulometricalJy at
about 50 milliamp and follow the potential comimiously with a digital voltmeter
or a pH meter provided with a millivolt .scale ; allow about 10 seconds near the
end-poini belorc taking potential readings, since equilibrium doc.s not appear to
be established immediately.

XIH, 14. CHLORIDE. URO.MIDE. AND IODIDE Discussion.

Mercuryd) ions can be generated at 100 per cent efliciency from
mcrcury-coatcd gold or from mercury pool anodes, and employed for
the coulomctric titration ol halides. The cnd-poinl is conveniently determined
potcntiomctrically. In titrations of chloride ion, the addition of methanol (up to
70-80 per cent) is desirable in order to reduce the solubility of the mercury(I)
chloride.

The standard potentials (r.v. N.M.E.) of the fundamental couples involving
uneompicxed mercury(I) and mercury(II) ions are;

Hg,-*" -f2t’ = 2Hg; - 4-0.80 volt

Hg-'^ -(-2e = Hg: - 4-0.88 volt

2Hg-" +2e = Hg,= ^ = -{-0.91 volt

560

COULOMETRY Xni, 14

The oxidation of Hg to Hg 2 ^'^ requires a smaller (less oxidising) potential than to
; mercury(I) ions are the main product when a mercury electrode is
subjected to anodic polarisation in a non-complexing medium. From a
stoichiometric standpoint it matters not whether oxidation of a mercury anode
produces the mercury(I) or mercury(II) salt of a given anion, because the same
quantity of electricity per mol of the anion is involved in either ease: thus the
same number of coulombs per mol of the anion are required to form either
Hg2Cl2 orHgCl2.

Apparatus. The apparatus is similar to that described in Section XIII, 9.
The generator anode A now consists of a mercury pool, 0.5-1 cm deep, at the
bottom of the cell; electrical connection is made by means of a platinum wire
sealed through glass tubing and dipping into the mercury. For titrations of
chloride and bromide, the mercury pool generator anode serves also as the
indicator electrode and is used in conjunction with a saturated calomel reference
electrode; the latter is conneeted to the cell through a saturated potassium nitrate
salt bridge. For titrations of iodide, the indicator electrode consists of a silver rod
fitted through glass tubing and held by the cover of the cell. During the titration
the contents of the electrolysis cell are stirred vigorously with a magnetic stirrer;
the stirrer bar floats on the surface of the mercury pool anode.

Reagents. Supporting electrolyte. For chloride and bromide,
use 0.5M-perchloric acid. For iodide, use O.lM-perchloric acid plus 0.4Af-
potassium nitrate. It is recommended that a stock solution of about five times the
above concentrations be prepared {2.5A/-perchloric acid for chloride and
bromide; 0.5M-perchloric acid + 2.0A/-potassium nitrate for iodide}, and
dilution be effected in the cell according to the volume of test solution used. The
reagents must be chloride-free.

Catlwlyte. The electrolyte in the isolated cathode compartment may be either
the same supporting electrolyte as in the cell or O.lM-sulphuric acid: the
formation of mercury(I) sulphate causes no difficulty.

Chloride. Experience in this determination may be obtained by the
titration of, say, carefully standardised ca. 0.005A/-hydrochloric acid.

Pipette 5.00 or 10.00 cm^ of the hydrochloric acid into the cell, add 35-40-cm^
of methanol and 10 cm^ of the stock solution of the supporting electrolyte. Fill
the isolated cathode compartment with supporting electrolyte of the same
concentration as that in the main body of the solution or with O.lM-sulphuric
acid ; the level of the liquid must be kept above that in the titration cell. Note the
counter reading, stir magnetically, and commence the electrolysis at about 50
milliamp. Stop the generating current periodically, record the counter reading,
and observe the potential between the mercury pool and the S.C.E. Plot a
potential-counter reading curve and evaluate the equivalence point from the first
or second differential graph. The approach of the equivalence point is readily
etected in practice : successive small increments of 0.05 or 0. 1 counter unit result
in a relatively large change of potential {ca. 30 millivolts per 0. 1 counter unit).

A R 0.0 IM solution of potassium bromide, prepared from the

• • salt previously dried at 1 1 0 °C, is suitable for practice in this determination,
e experimental details are similar to those given above for Chloride except that
brn^^a added. The titration cell may contain 10.00 cm^ of the

solution, 30 cm^ of water, and lOcm^ of the stock solution of

supporting electrolyte.

Iodide. A O.OIM solution of potassium iodide, prepared from the dry A.R.

561

XUI, 15/16 QUANTITATIVE INORGANIC ANALYSIS

salt with boiled-oul water, is suitable for practice in this detennination. The
experimental details are similar to those given for Broimle, except that the
indicator electrode consists ofa silver rod imntersed in the solution. The titration
cell may be charged with lO.OOcm-* of the iodide solution, 30 cm^ of water, and
lOcnf* of the stock solution of perchloric acid + potassium nitrate. In the
neighbourhood of the equivalence point it is necessary to allow at least 30-60
seconds to elap.se before steady potentials arc established,

Xlll, 15. BROMIDE AND IODIDE, Discussion. Silver ion can be
electrogencraied with IdO per cent clliciency at a silver anode and can be applied
to precipitation titrations. The end-points can be determined potentiomctrically
or, less accurately, visually with ad.^orpIion indicators (in halide determinations;
eosin for bromide and iodide; dichlorolluoresccin for chloride). The insoluble
silver salt deposits on the silver anode and the solution remains clear until the
residua! concentration of halide ion becomes so small that its rate of transfer to
the anode is smaller than the rate at which silver ion is generated ; thenceforth the
silver ion produced at the anotic diffuses into the solution and the precipitation
occurs in the solution.

The supporting electrolyte may be 0.5il/-potassium nitrate for bromide
and iodide; for chloride, 0.5A/-poiassium nitrate in 25-30 percent ethanol must
be iLsed becau.se of the appreciable solubility of silver chloride in water.

Apparatiw. Use the apparatus of SecUon .\in,9. The generator anode is of
puresiUcT foil (3 x 3 cm): the cathode in the isolated compartment is a platinum
foil (3x3cm) bent into a half-cylinder. For the potentiomelric end-point
detection, use a short length of platinum or silver wire as the indicator electrode;
the electrical connection to the saturated calomel reference electrode is made by
means of an agar-pota.ssium nitrate bridge.

ReagenLs. Siippurtini; clectrolyic. Prepare 0.5;\/-pola.ssiinn nitrate from
the A.R. salt.

PoHissiwn bromiilc solution, ca. 0.025M. Weigh accurately the appropriate
amount of dry A.R. potassium bromide.

Procedure. Broniidc. Place 40 cm^ of the supporting electrolyte in the cell
and add 5.00, 10,00, or IS.OOcni^ of the potassium bromide solution. Charge the
isolated cathode compartment with ().5,tf-potassium nitrate. Pass a current of 30
milliamp, stir vigorously, and measure the potential of the indicator electrode:
the potential change at the end-point is about 200 millivolts.

Iodide. Proceed as for Bromide.

XIII, 16. Ill RATION OF ACIDS. Generol discussion. The limiting
reactions in aqueous solution at platinum electrodes are:

21-1,0 ^O; +4c(anode)

2H,0 + 2c M, -f- 201 1 " (cathode)

consequently anodic clectro-gencrulion of hydrogen ion for the titration of bases
and cathodic elcclro-gcncration of hydroxide ion* for the titmtion of acids is

Direct reduction ol hydrogen ion (2il' mjy occur with Mn.ill eutrenl densities at a

platinum cathode. It is nnniaicnal in .j stoichiumcirie sense u liclher the titration proceeds indirectly
by the electro-generated hydroxide ion or directly by the reduction of hydrogen ion; the current
etiicicney remains at lOO per cent up to scry large current densities.

562

COULOMETRY XIII, 16

readily accomplished. One of the many advantages of coulometric titration of
acids is that difficulties associated with the presence of carbon dioxide in the test
solution or of carbonate in the standard titrant base are easily avoided: carbon
dioxide can be removed completely by passing nitrogen or carbon dioxide-free
air through the original acid solution before the titration is commenced. The
presence of any substance that is reduced more easily than hydrogen ion or water
at a platinum cathode, or which is oxidised more easily than water at a platinum
anode, will, of course, interfere.

When internal generation is used in association with a platinum auxiliary
electrode the latter must be placed in a separate compartment (see Fig. XIII, 10) ;
contact between the auxiliary electrode compartment and the sample solution is
made through some sort of a diaphragm, e.g., a tube with a sintered glass disc or
an agar-salt bridge. For the titration of acids a silver anode may be used in
combination with a platinum cathode in presence of bromide ions ; the silver
electrode is placed inside a straight tube closed by a sintered disc at its lower end
and this can be inserted directly into the test solution. A bromide ion
concentration of about 0.05M is satisfactory.

A. With isolated platinum auxiliary generating electrode.
Apparatus. Use the cell (ca. 150cm^ capacity) shown in Fig. XIII, 6

Section XIII, 6), but equipped as pre-
y Stirring bar sented diagrammatically in Fig. XIII, 10.

The working electrode consists of a
Agar-KCl W ^ ^ „ platinum foil (4x2 cm); the auxiliary

bridge I O \ ■ electrode is a small platinum sheet

N / ^ VA — Generator (1x1 cm) immersed in a small beaker

Y I electrode connected to the cell by means of an

\ O C\ J inverted U-tube salt bridge containing 3

\ T cent agar gel in saturated potassium

\A chloride. The glass electrode and the

Srode saturated calomel reference electrode

are those supplied with commercial pH
Fig. xm 10 meters. Efficient stirring is provided by a

magnetic stirrer.

Reagents. Supporting electrolyte. O.lAf-sodium chloride solution.

Catholyte. This consists of O.lAZ-sodium chloride solution to which a little
dilute sodium hydroxide solution is added.

Hydrochloric acid, O.OIM and O.OOIM. Prepare with boiled-out water using
A.R. hydrochloric acid, and standardise.

isolated

Stirring bar

Agar-KCl
bridge I

-Generator

electrode

Auxiliary

electrode

Glass—'

electrode

Fig.xin,io

Procedure. Place 50 cm^ of the supporting electrolyte in the coulometric
cell, and pass nitrogen through the solution until a pH of 7.0 is attained:
thenceforth pass nitrogen over the surface of the solution. Pipette lO.OOcm^ of
ffie acid into the cell. Adjust the current to a suitable value (40 or 20 milliarrip).
Turn on the current and read the counter of the integrating motor
simultaneously: stop the titration when the equivalence point pH (7.00) is
reached. ■ •

B. With silver auxiliary electrode. Apparatus. The titration cell is shown
lagrammatically in Fig. XIII, 1 1 , but note that the silver anode is placed inside a
g ass tube (not shown) with a sintered disc at the lower end — see Fig. XIII, 5. It
consists of a 100-1 50-cm^ rimless beaker. The cork or plastic co ver has holes for
W inlet and outlet tubes for nitrogen, (ii) platinum, cathode and silver anode, and

563

XllI, 17/18 QUANTITATIVE INORGANIC ANALYSIS

Glass clccifodc

Pt calhodc -

S.C.E-

Ag ancxlc

FiS>.XlIl, 11

(iii) a glass electrode and a saturated calomel
reference electrode such as are supplied with
commercial pll meters. If the S.C.E. cannot be
accommodated conveniently in the cell it may be
placed in a small beaker of saturated potassium
chloride solution and connected to the test solution
by a Li'-tube salt bridge containing saturated
potassium chloride solution in 3 percent agar. Both
the platinum cathode and the silver anode consist
of stout wires coiled into helices. The silver anode
may be used repeatedly before the silver bromide
coatingbecomesso thick that it must beremoved—
about thirty successive titrations of 0.1 mcq.
samples at 20 milliamp. Wlicn finally necessary, the
silver bromide coating may be removed by
dissolution in potassium cyanide solution.
electrolyte. Prepare a 0.05 .t /-sodium bromide

Reagent. Xuppartuig
solution using the A.R. salt.

Pruceditre. Place 50cm' of the supporlmg electrolyte in the beaker and
add some of the same solution to the tube carrying the silver electrode so that the
liquid level in this tube is just above the beaker. Pa.ss nitrogen into the solu-
tion until the pH is 7,0. Pipette lO.OOcrn' of either 0.01. 1/- or 0.001. l/-
hydrochloric acid into tire cell. Continue the passage of nitrogen. Proceed with
the titration as described under A above.

Several successive iamplcs may be titrated without renewing the supporting
electrolyte.

Note. The above techniques arc generally applicable to many other acids,
both strong and weak. The only limitation is that the anion must not be reducible
at the platinum cathode and must not react in any way with the silver anode or
with silver bromide (e.g., by eomple.xation).

The coulometric determination of acids has been extensively studied by Bishop
and co-workers ( Ref. 4 and subsequent papers).

XIII, 17. TITRATION OF BASES. Discussion. Whenabaseistitratedwith
electro-generated hydrogen ion at a platinum anode ( 2 H 2 O O^T 4H’'t 4e)a
platinum auxiliary catijode is u.sed and must be separated from the test solution
by placing it in a separate compartment. The apparatus described under
Electrolyikolly Generated Hydroxide Ion (Section Xlll, 16, .A) may be employed;
the electrodes are. ofeourse, rever.sed.

Reagent. Siipportiny electrolyte. Prepare a t).2.U-sodiuni sulphate
.solution using the A.R. salt.

Procedure. Experience in this titration may be acquired by titration of,
say, 5.00 cm' of accurately standardised O.OlA'-sodium hydroxide solution. Use
50 cm' of. supporting electrolyte and a current of 30 milliamp.

XIII, 18. References

1 . E. Bishop and P. II. llitcheoek (1973). •Poicniioslalie Coulometric Determination of
Vanadium, Vanadium -Manganese and Vanadium-Iron Mixtures'. Analyst, 98, 574.

564

COULOMETRY xni,19

2. I. R. Juniper (.1974). ‘A Solid-state Potentiostat for Controlled Cathode-potential
Electrolysis’. Analyst, 99, 58.

3. E. Bishop and P. H. Hitchcock (1973). ‘Mass and Charge Transfer Kinetics and
Coulometric Current ElBciencies’. Analyst, 98, 470.

4. E. Bishop and M. Riley (1973). ‘Precise Coulometric Determination of Acids in Cells
without Liquid Junctions’. Part I. Analyst, 98, 305.

5. J. A. Pike and G. C. Goode (1967). ‘Precise Constant-current Coulometer’. Anal.
Chimica Acta, 29,1.

6. G. Marinenko and J. K. Taylor (1968). ‘Electrochemical Equivalents of Benzoic and
Oxalic Acid’. Anal. Chem., 40, 1645.

7. J. J. Lingane (1958). Electroanalytical Chemistry. 2nd edn. New York; Interscience.

8. V. J. Jennings, A. Dodson and A. Harrison (1974). ‘Coulometric Micro-titration of
Arsenic(III) and Isoniazid Using a Vitreous Carbon Generating- Electrode’. Analyst,
99, 145.

Xin, 19. Selected bibliography

1. D. R. Browning (1969). Electrometric Methods. London; McGraw-Hill.

2. D. G. Davis (1972). ‘Electroanalysis and Coulometric Analysis’ (review article). Anal.
Chemistry, 44, 79R.

3. P. Delahay (1957). Instrumental Analysis. New York; The Macmillan Co.

4. J. G. Dick (1973). Analytical Chemistry. New York; McGraw-Hill Inc.

5. G. W. Ewing (1975). Instrumental Methods of Chemical Analysis. 4th edn. New York;
McGraw-Hill Book Co.

6. B. Fleet and R. D. Jee (1973). Electrochemistry. Vol. 3. Specialist Periodical Report.
London; The Chemical Society.

7. G. G. Guilbault and L. G. Hargis (1970). Instrumental Analysis Manual. New York ;
Marcel Dekker Inc.

8. I, M. Kolthoff and P. J. Elving (1963). Treatise on Analytical Chemistry. Part I. Vol. 4.
New York; Wiley-Interscience.

9. J. J. Lingane (1958). Electroanalytical Chemistry. 2nd edn. New York; Interscience.

10. L. Meites and H. C. Thomas (1958). Advanced Analytical Chemistry. New York;
McGraw-Hill Book Co.

11. W. F. Pickering (1971). Modern Analytical Chemistry. New York; Marcel Dekker
Inc.

12. H. A. Strobel (1973). Chemical Instrumentation. A Systematic Approach to
Instrumental Analysis. 2nd edn. Reading, Mass., Addison-Wesley Publishing Co.

13. T. S. West (1973). Analytical Chemistry. Part 2. (MTP Series). London; Butterworth
and Co.

14. H. H, Willard, L. L. Merritt and J. A. Dean (1974). Instrumental Methods of Analysis.
5th edn. New York; Van Nostrand.

15. C. Woodward and H. N. Redman (1973). High-precision Titrimetry. Analytical
Sciences Monograph. No. 1 . London ; Society for Analytical Chemistry.

16. D. A. Skoog and D. M. West (1971). Principles of Instrumental Analysis. New York;
Holt, Rinehart and Winston Inc.

17. C. L. Wilson and D. W. Wilson (197 1). Comprehensive Analytical Chemistry. Vol. 2B.
London; Elsevier.

565

CHAPTER XIV POTEWTIOIVIETRY

XIV, 1. INTRODUCnON. As shown in Seciion 11, 20, when a metai M is
immersed in a soiuiion containing its own ions M* *■, then an electrode potential
is established, the value of whiclt is given by the jNcrast equation

£ = £^* +(KT/nf iliuv.

where is a constant, the standard electrode potential of the metal. £can be
measured by combining the electrode with a reference electrode (commonly a
saturated calomel electrode: see Section XIV, 3), and measuring ihec.m.f. of the
resultant cell. It follows that knowing the potential £, of the reference electrode,
sve can deduce the value of the electrode potential £, and provided the standard
electrode potential E'-^ of the given metal is known, we can then proceed to
calculate the metal ion activity u..,.. in the .solution. For a dilute solution the
measured ionic activity will be virtually tlic same as the ionic concentration, and
for stronger solutions, given the value of the activity cocllkicai, we can convert
the measured ionic activity into the corresponding concentration.

This procedure of using a single niea.surcment of electrode potential to
determine the concentration of an ionic species in solution is referred to as direct
potciitionietry. The electrode whose potential is dependent upon the
concentration of the ion to be determined is termed the indicator electrode, and
when, as in the case above, the ion to be detcrmiitcd is directly involved in the
electrode reaction, we are said to be dealing with an ekctriuh; o/ilie first kind.

U is also possible in appropriate cases to measure by direct potcnliomeir>’ the
concentration of an ion winch is not directly concerned in the electrode reaction.
This inv olves the tise of an i'kc trade of the second kind, an example of which is the
silver-silver chlorklc electrode which is formed by coating a silver vvite with silver
chloride; this electrode can be used to measure the concentration of chloride ions
in soiuiion.

The silver wire can be regarded as a silver electrode with a potential given by
the Nernst equation as

£ = £t, + (RT/»i£)lnu,,.

The silver ions involved are derived from the silverchloride. and by the solubility
product principle (Section II, 8 ), the activity of these ions will be governed by the
chloride ion activity

566

POTENTIOMETRY XIV, 1

Hence the electrode potential can be expressed as . '

E = E%+(RTfnF)\nK,-(RT/riF)lnaa- .

and is clearly governed by the activity of the chloride ions, so that the value of the
latter can be deduced from the measured electrode potential.

In the Nemst equation the term RTjnF involves known constants, and
introducing the factor for converting natural logarithms to logarithms to base
10, the term has a value at a temperature of 25 °C of 0.0591 V when n is equal to 1.
Hence, for a univalent metal, a tenfold change in ionic activity will alter the
electrode potential by about 60 millivolts, whilst if the metal is bivalent, a similar
change in activity . will alter the electrode potential by approximately 30
millivolts, and it follows that to achieve an accuracy of 1 per cent in the value
determined for the ionic concentration by direct potentiometry, the electrode
potential must be capable of measurement to within 0.26 mV for the univalent
metal, and to within 0.13 mV for the bivalent metal.

An element of uncertainty is introduced into the e.m.f. measurement by the
liquid junction potential which is established at the interface between the two
solutions, one pertaining to the reference electrode and the other to the indicator
electrode. This liquid junction potential can be largely eliminated however if one
solution contains a high concentration of potassium chloride or of ammonium
nitrate; electrolytes in which the ionic conductivities of the cation and the anion
have very similar values.

One way of overcoming the liquid junction potential problem is to replace the
reference electrode by an electrode composed of a solution containing the same
cation as in the solution under test, but at a known concentration, together with a
rod of the same metal as that used in the indicator electrode: in other words we
set up a concentration cell (Section U, 21). The activity of the metal ion in the
solution under test is given by

£„i. = [RTInF)\n

(actim'ty)„„know.

As a further refinement of this procedure, provided that we start with a solution
containing a known ionic concentration which is greater than that in the solution
under measurement, then by a process of accurate dilution of the standard
solution, we can adjust its concentration to be the same as that in the solution
under test. This process will be accompanied by a gradual fall in the e.m.f. of the
concentration cell, and when the two solutions have the same concentration the
cell e.m.f. will be zero ; this procedure is termed null point potentiometry.

In view of the problems referred to above in connection with direct
potentiometry, much attention has been directed to the procedure of
potentiometric titration as an analytical method. As the name implies, it is a
titration procedure in which potentiometric measurements are carried out in
T point. In this procedure we are concerned with changes in

e ectrode potential rather than in an accurate value for the electrode potential
w a given solution, and under these circumstances the effect of the liquid
junction potential may be ignored. In such a titration, the change in cell e.m.f.
occurs inost rapidly in the neighbourhood of the end point, and as will be
xp ained later (SectiomXIV, 23), various methods can be used to ascertain the
of th potential change is at a maximum: this is the end point

567

XIV, 2 QUANTITATIVE INORGANIC ANALYSIS

In the present Chapter consideration will be given to various types of indicator
and reference electrodes, to the procedures and instrumentation (poten-
tiometers) for measuringcell e.m.f., to sonic selected examples of determinations
carried out by direct potentiometry, and to some typical examples of
potentiomelric titrations.

Reference electrodes

XIV, 2. THE HYDROGEN ELECIRODE. All electrode potentials arc
quoted witli reference to the standard hydrogen electrode (Section 11, 20). and
hence this must be regarded as the primary reference electrode. A typical
hydrogen electrode has already been described (Section II, 20), and the electrode
shown in Fig. II, 2 is the Hildehranil bdl-type electrode. The platinum electrode is
surrounded by an outer tube into which hydrogen enters through a side inlet,
escaping at the bottom through the test solution. There arc several small holes
near the bottom of the bell: when the speed of the gas is suitably adjusted, the
hydrogen escapes through the small openings only. Becau.se of the periodic
formation of bubbles, the level of the liquid inside the tube fluctuates, and a part
of the foil is alternately c.xposed to the solution and to hydrogen. The lower end
of tlic foil is continuously immersed in the solution to avoid interruption of the
electric current. It siunild be noted that although in Fig. II, 2 an open vessel is
shown, in practice the electrode will be used in a stoppered llask with a suitable
exit for the hydrogen, so that an oxygen-free atmosphere can be nuiintaincd in
the flask.

The Lindsey hydrogen electrode, illustrated in Fig. XIV, 1 has many valuable
features and utilises 5- 7cnri of the test solution. The introduction and the
removal of the tcsl solution is simple, and rapid saturation of the platinum is

readily attained. Tlic hydrogen-outlet trap is
at right angles to the plane of the paper; and
not in the same plane as indicated in the figure.
The funnel limb serves for filling and washing
out the vessel, and also supplies the con-
nection to the reference electrode. Tlie
hydrogen stream is admitted through the left-
hand tube and is adjusted to produce a
pulsating movement up and down the plati-
num electrode.

POTENTIOMETRY XIV, 3

obtained from two accumulators connected to a suitable sliding resistance; the
current is adjusted to produce a moderate evolution of hydrogen, and the process
is complete in about 2 minutes. It is important that only a thin, jet-black deposit
be made; thick deposits lead to unsatisfactory hydrogen electrodes. After
platinising, the electrode must be freed from traces of chlorine: it is washed
thoroughly with water, electrolysed in ca. 0.25M-sulphuric acid as cathode for
about 30 minutes, and again well washed with water. Hydrogen electrodes
should be, stored in distilled water; they should never be touched with the fingers.
It is advisable to have two hydrogen electrodes so that the readings obtained with
one can be periodically checked against the other.

The most convenient source of hydrogen is the compressed gas, sold in
cylinders ; a steady stream of hydrogen gas may be readily obtained by means of a
reducing valve. The gas may be passed through all-glass wash bottles containing
respectively 0.2M-potassium permanganate solution, alkaline pyrogallol
solution (1-2 g of pyrogallol in ca. 35 cm^ of 4M-sodium hydroxide solution),
dilute sulphuric acid (ca. 0.05Af ; to neutralise alkali which might splash over),
and distilled water, before reaching the electrode.

The alkaline pyrogallol serves to remove any traces of oxygen from the gas :
this is most important, for otherwise it is difficult to establish a steady electrode
potential owing to interaction between hydrogen and oxygen on the platinised
surface of the electrode. An alternative procedure for removing oxygen is to pass
the gas over heated platinised asbestos, but if this method is used, particular care
must be taken to ensure that the gas is properly cooled and saturated with water
vapour before admission to the electrode. Connections in the gas supply line
should preferably be made with polythene tubing : rubber tubing should not be
used unless it has been treated with hot concentrated sodium hydroxide solution
and then thoroughly washed in order to remove traces of sulphur compounds
which might ‘poison’ the electrode.

Of the two electrodes shown, the Lindsey pattern is particularly suited for use
as a reference electrode, whilst the alternative Hildebrand electrode has
advantages if the electrode is to function as an indicator electrode (Section XIV,
6), and especially for potentiometric titrations.

Although the hydrogen electrode is the primary reference electrode, in
practise, subsidiary standard electrodes which can be kept permanently set up
and which are therefore available for immediate use are preferred for most
purposes, thus obviating the careful setting up (including gas purification) which
IS required in order to establish a satisfactory hydrogen electrode. When used as a
standard electrode, the hydrogen electrode operates in a solution containing
hydrogen ions at constant (unit) activity based usually on hydrochloric acid, and
the hydrogen gas must be at one atmosphere pressure ; the effect of change in gas
pressure is discussed in Ref. 1 .

XIV, 3. THE CALOMEL ELECTRODE. The most widely used reference
e ectrode, due to its ease of preparation and constancy of potential, is the calomel
^ calomel half-cell is one in which mercury and calomel (mercury(I)
c londe) are covered with potassium chloride solution of definite concentration ;
IS may be 0. 1 A, 1 .ON, 3.5iV, or saturated. The potassium chloride solution must
e saturated with the calomel. The potentials of the O.LV, l.OA^, and saturated
calomel electrodes at 25 °C relative to the normal hydrogen electrode are 0.337 1,
•2646, and 0,2458 volt respectively.

569

XIV, 3 QUANTITATIVE INORGANIC ANALYSIS

Various forms of the calomel electrode arc illustrated in Fig. XIV, 2; O.IN, N,
or saturated potassium chloride may be used, but the last-named is generally
preferred for outine work. One of these (n), will be described in detail; the others
will then be self-evident. It consists of a glass ve.ssel provided with a bent side tube
A and another side tube B, over the end of which a piece of rubber tubing is
placed which can be closed by :i spring or screw clip. Electrical connection with
the electrode is made by means of a platinum wire, sealed through a glass tube C;
the latter contains a little pure mercury' into which an amalgamated copper wire

Fig. XIV, 2

dips. To set up the electrode, a saturated solution of analytically pure potassium
chloride containing some of the solid salt is first prepared. Pure mercury to a
depth of 0.5- 1 cm is placed in the bottom of the dry electrode vessel ; the mercury
is then covered with a layer of calomel paste D. The latter is prepared by rubbing
pure calomel, mercury, and saturated potassium chloride solution together in a
glass mortar; the supernatant liquid is poured olf and the rubbing process
repeated twice with fresh quantities of saturated potassium chloride solution.
The rubber bung carrying the glass tube and platinum wire is then inserted, care

being taken that the platinum wire dips into
the mercury. The vessel is then filled with a
saturated solution of potassium chloride
{previously saturated with calomel by
shaking with the solid salt) by drawing in
the solution through the bent tube A. and
then closing the rubber tube B with a clip.
The electrode is then ready for use. In
electrode (b), the siphon tube or salt bridge
may be filled with a jelly of 3 percent agar in
saturated potassium chloride solution. The
electrode (c) is suitable for precision work;
omt paac three-way stopcock for flushing

away the contaminated potassium chloride
after it has been employed in a titration.
Compact calomel electrodes ate available
commercially (sec Fig. XIV, 3).

For special purposes, modifleations of
the calomel electrode may be preferred.

570

_ POTENTIOMETRY XIV, 4/5

Thus if it is necessary to avoid the presence of potassium ions (see for example
the determination of potassium by amperometric titration— Section XVII, 9),
the electrode may be prepared with sodium chloride. solution replacing the
potassium chloride. In some cases the presence of chloride ions may be inimical
and a mercury(I) sulphate electrode may then be used : this is prepared in similar
manner to a calomel electrode using mercury(I) sulphate and potassium or
sodium sulphate solution.

XIV, 4. THE SILVER-SILVER CHLORIDE ELECTRODE. This elec-
trode is perhaps next in importance to the calomel electrode as a reference
electrode. It consists of a silver wire or a silver-plated platinum wire, coated
electrolytically with a thin layer of silver chloride, dipping into a potassium
chloride solution of known concentration. The potentials of the O.IM and
saturated silver-silver chloride electrodes at 25 °C with respect to the normal (or
standard) hydrogen electrode are 0.290 and 0.199 volt respectively. In certain
circumstances electrodes other than those described above may serve as reference
electrodes, but generally these are of limited application.

Indicator electrodes

XIV, 5. GENERAL DISCUSSION. As already stated, the indicator
electrode of a cell is one whose potential is dependent upon the activity (and
therefore the concentration) of a particular ionic species whose concentration is
to be determined. In direct potentioraetry or the potentiometric titration of a
metal ion, a simple indicator electrode will usually consist of a carefully cleaned
rod or wire of the appropriate metal : it is most important that the surface of the
metal to be dipped into the solution is free from oxide films or any corrosion
products. In some cases a more satisfactory electrode can be prepared by using a
platinum wire which has been coated with a thin film of the appropriate metal by
electro-deposition.

When hydrogen ions are involved, a hydrogen electrode can obviously be used
as indicator electrode, but its function can also be performed by other electrodes,
foremost amongst which is the glass electrode. This is an example of a membrane
electrode in which the potential developed between the surface of a glass
membrane and a solution is a linear function of the pH of the solution, and so can
be used to measure the hydrogen ion concentration of the solution. Since the
glass membrane contains alkali metal ions, it is also possible to develop glass
electrodes which can be used to determine the concentration of these ions in
solution, and from this development (which is based upon an ion exchange
mechanism), a whole range of membrane electrodes have evolved based upon
both solid state and liquid membrane ion exchange materials : these electrodes
constitute the important series of ion sensitive electrodes which are now available
for many different ions (Sections XIV, 9-12).

Indicator electrodes for anions may take the form of a gas electrode (e.g.,
oxygen electrode for OH' ; chlorine electrode for Cl'), but in many instances
wmist of an appropriate electrode of the second kind ; thus as shown in Section
^V, 1, the potential of a silver-silver choride electrode is governed by the
chloride ion activity of the solution. Ion sensitive electrodes are also available
‘Or many anions.

The indicator electrode employed in a potentiometric titration will of course be

571

XIV, 6/7 QUANTITATIVE INORGANIC ANALYSIS

dependent upon the type of reaction which is under investigation. Thus, for an
acid-base titration, the indicator electrode may be a hydrogen electrode or some
other hydrogen-ion responsive electrode (Sections XIV, 7-8) ; for a precipitation
titration (halide with silver nitrate, or silver with chloride) a silver electrode will
be used, and fora rcdo.x titration (e.g.. iron(ll) w'iihdichroinate) a plain platinum
wire is used as the rcdo.x electrode.

XIV, 6. THE HYDROGEN ELECTRODE. In addition to its function as a
standard electrode, the hydrogen electrode can be used as indicator electrode to
measure the hydrogen ion concentration or the pH of solutions and can also be
employed for potentiometric acid-base titrations.

The construction and operation of such electrodes have already been described
(Section XIV, 2). but it must be noted that the hydrogen electrode cannot be used
in solutions containing oxidising agents, e.g., permanganate, nitrate, ccrium(lV)
and iron(lll) ions, or of other substance.s capable of reduction, such as
unsaturated organic compounds, or in the presence of sulphides, compounds of
arsenic, etc. (catalytic poisons) which destroy the catalytic property of platinum
black. It is also unsatisfactory in the presence of salts of the noble metals, e.g.,
copper, silver, and gold, and also in solutions containing lead, cadmium, and
thalliumd ) salts. There are many other electrodes which are more convenient to
use in the range in which they are applicable. Some of these will be described
below.

Mention must, however, be made of the advantages of the hydrogen electrode:
(1) it is a fundamental electrode to which all measurements of pH are ultimately
referred : (3) it cun be applied over the entire pi I range ; and (3) it e.xhibits no salt
error.

Just as for reference electrode purposes the hydrogen electrode is more
conveniently replaced by alternatives such as the calomel or the silver-silver
chloride electrode, so too alternative electrode.', arc preferred to the hydrogen
electrode as an indicator electrode. Amongst these alternative hydrogen-ion
responsive electrodes may be mentioned (i) the quinhydrone electrode (of
historic interest and now rarely used for analytical purposes), (ii) the antimony
electrode (of limited application in the analytical held, but of .some industrial
importance on account of its simple nature and robust character), and (iii) the
glass electrode: the latter has virtually superseded ail other electrodes for the
measurement of hydrogen ion concentration.

XIV, 7. THE .VNTI.MONY ELECTRODE. The so-called 'antimony elec-
trode’ is really an antimony-antimony trioxide electrode. The electrode reaction
is:

SbjOj (.V) + 6 1 H -r 6c ^ 2Sb (.V) + 3 H ,0

and the potential at 25 "C is theoretically given by:

,, _o 0.0591 . 1

^ = ^S,o,„sb log-;.- = £S_o,.sh- 0.0591 pH

the activities of the solid antimony and antimony trioxide, and of the water, being
taken as unity. In practice, it is found that the pH response of the Sb.SbjOi
electrode is roughly given by the above equation, but the exact limits of the
validity of this relation arc uncertain.

572

POTENTIOMETRY XIV, 8

The electrode is generally prepared by casting a stick of antimony in the
presence of air: sufficient oxidation occurs in this way to render further addition
of oxide unnecessary. A wire is attached to one end of the antimony rod, while the
other end is inserted into the experimental solution: the potential is then
measured against a convenient reference electrode. As the potentials dilfer from
one electrode to another, it is necessary to standardise each antimony electrode
by means of solutions of known pH (buffer solutions) and also under the same
experimental conditions to which it will be subjected in use; for example, in the
presence or absence of oxygen, etc. The most useful pH range is 2-8.

The antimony electrode cannot be applied: (a) in the presence of strong
oxidising agents or of complexing reagents (such as tartrates and organic
hydroxy acids) ; (b) in solutions with a pH lower than 3, since the oxide then
beomes appreciably soluble; and (c) in the presence of metals more noble than
antimony. The electrode is not readily poisoned, is simple to use (no reagents are
usually required), and is rugged; it has therefore found application for the
continuous recording or control of pH in conditions where it is applicable.

XTV, 8. THE GLASS ELECTRODE. The glass electrode is the most widely
used hydrogen-ion responsive electrode, and its use is dependent upon the fact
that when a glass membrane is immersed in a solution, a potential is developed
which is a linear function of the hydrogen ion concentration of the solution. The
basic arrangement of a glass electrode is shown in Fig. XIV, 4 (a ) ; the bulb B is
immersed in the solution of which it is required to measure the hydrogen ion

concentration, and the electrical circuit is
completed by filling the bulb with a
solution of hydrochloric acid (usually
0. 1 M), and inserting a silver-silver chloride
electrode. Provided that the internal hy-
drochloric acid solution is maintained at
constant concentration, the potential of the
silver-silver chloride electrode inserted
into it will be constant, and so too will the
potential between the hydrochloric acid
solution and the inner surface of the glass
bulb. Hence the only potential which can
vary is that existing between the outer
surface of the glass bulb and the test
solution into which it is immersed, and so
the overall potential of the electrode is
governed by the hydrogen ion concen-
tration of the test solution. To ensure that
the concentration of the inner hydrochloric
acid solution remains constant, the upper
end of the electrode must be sealed, thus
giving the typical glass electrode depicted
in Fig. XIV, 4(b).

The nature of the glass used for con-
struction of the glass electrode is very
important. Hard glasses- of the Pyrex type
are not suitable, and for many years a

573

XIV, 8 QUANTITATIVE INORGANIC ANALYSIS

lime-soda glass (Corning 015) of the approximate composition SiO, 72 per cent.
Na^O 22 per cent, CaO 6 per cent was universally used for the manufacture of
glass electrodes. Such electrodes were extremely satisfactory over the pH range
1-9, but in solutions of higher alkalinity the electrode was subject to an ‘alkaline
error’ and tended to give low values for the pH. The error increased with the
concentration of alkali metal ions in solution, and for e.xample at pH 12 in the
presence of sodium ions, the error varied from - !.0pH((Na’’j = l.U) to
-0.4pH([Na’] = O.LU): the errors were .smaller in .solutions containing
lithium, potassium, barium or calcium ions. Attempts were therefore made to
discover glasses which would give electrodes free from this alkaline error, and it
was found that the required result could be achieved by replacing most or all of
the sodium content of the gla.ss by lithium, and an electrode constructed of a glass
having the composition SiO, 63 percent, Li,02S percent. 05,02 percent, BaO
4 per cent, La20j 3 per cent has an error of only ~0.12pH at pH 12.8 in the
presence of sodium ions at a concentration of 2.U. Lithium-based glasses are now
exclusively used for hydrogen-ion respon.sive glass electrodes.

To measure the hydrogen ion concentration of a solution the gla.s.s electrode
must be combined with a reference electrode, for which purpose the saturated
calomel electrode i-s most commonly u.sed, thus giving the cell ;

Ag.AgClH) I HCKO.l.U) ! Gla-ss | Test .solution || KCI(.5at’d),HgiCl,(.r) i Hg.

Owing to the high resistance of the glass membrane, a simple potentiometer
(Section XIV, 13) cannot be employed for measuring the cell c.m,f. and
.specialised instrumentation (Section .XIV, 14) must be used. The c.m.f. of theccll
may be expressed by the equation :

£= A'4-(;Tr//0lnr/i,.

or at a temperature of 25 by the expression ;

E= A:-^0.0591 pH.

In these equations K is a constant partly dependent upon the nature of the glass
used in the conslruciion of the membrane, and partly upon the individual
character of each electrode; its value may vary slightly with time. This variation
of K with time is related to the existence of an asymmetry poieniuil in a glass
electrode which is determined by the dilfering responses of the inner and outer
surfaces of the ghiss bulb to changes in hydrogen ion activity ; this may originate
as a resuit of difl'ering conditions of strain in the two glass surfaces. Owing to the
asymmetry potential, if a glass electrode is inserted into a test solution which is in
fact identical with the internal hydrochloric acid .solution, then the electrode has
a small potential which is found to vary with time. On account of the existence of
thi.s a-syrnmetry potential of time-dependent magnitude, a constant value cannot
be assigned to K, and every glass electrode mu.st be standardised frequently by
placing in a solution of known hydrogen ion acti\ it y (a buifer solution).

So-called combination electrodes may be purchased in which the glass
electrode and the saturated calomel reference electrode ;irc combined into a
single unit, thus giving a more robust piece of equipment, and the convenience of
having to insert and support a single probe in the test solution instead of the two
separate components.

As will be apparent from the above discussion, the operation of a glass
electrode is related to the situations existing at the inner and outer surfaces of the

574

POTENTIOMETRY XIV, 9

glass membrane. Glass electrodes require soaking in water for some hours before
use and it is concluded that a hydrated layer is formed on the glass surface, inside
which an ion exchange process can take place. If the glass contains sodium, the
exchange process can be represented by the equilibrium

Hjtln "b hlUgiass ^ Hglass 'b N^oln ,

The concentration of the solution within the glass bulb is fixed, and hence on the
inner side of the bulb an equilibrium condition leading to a constant potential is
established. On the outside of the bulb, the potential developed will be dependent
upon the hydrogen- ion concentration of the solution in which the bulb is
immersed. Within the layer of ‘dry’ glass which exists between the inner and outer
hydrated layers, the conductivity is due to the interstitial migration of sodium
ions within the silicate lattice. For a detailed account of the theory of the glass
electrode a text book of electrochemistry should be consulted.

In view of the equilibrium shown in the equation above it is not surprising that
if the solution to be measured contains a high concentration of sodium ions, say a
sodium hydroxide solution, the pH determined is too low. Under these
conditions sodium ions from a solution pass into the hydrated layer in preference
to hydrogen ions, and consequently the measured e.m.f (and hence the pH) are
too low. This is the reason for the ‘alkaline error’ encountered with the glass
electrode constructed from a lime-soda glass. Likewise in strongly acid solutions
(hydrogen ion concentration in excess of \M), errors also arise but to a much
smaller degree; this effect is related to the fact that in the relatively concentrated
solutions involved, the activity of the water in the solution is reduced and this can
affect the hydrated layer of the electrode which is involved in the ion exchange
reaction.

The glass electrode can be used in the presence of strong oxidants and
reductants, in viscous media, and in the presence of proteins and similar
substances which seriously interfere with other electrodes. It can also be adapted
for measurements with small volumes of solutions. It may give erroneous results
when used with very poorly buffered solutions which are nearly neutral.

The glass electrode should be thoroughly washed with distilled water after
each measurement and then rinsed with several portions of the next test solution
before making the following measurement. The glass electrode should not be
allowed to become dry, except during long periods of storage: it will return to its
responsive condition when immersed in distilled water for at least twelve hours
prior to use.

Ion-sensitive electrodes

XIV, 9. ALKALI METAL ION-RESPONSIVE GLASS ELECTRODES.
As nientioned in Section XIV, 8, a glass electrode used for pH measurements, will
if it is constructed of a lime-soda glass, be subject to an ‘alkaline error’ which
sterns from the ion exchange equilibrium between hydrogen ions in solution and
sodium ions in the layer of hydrated glass. If the corhpbsition of the glass is
altered, theri so too is the position of equilibrium, and indeed, as already stated, if
the sodium in the glass is replaced by lithium, then the ‘alkaline error’ virtually
disappears.

If the preference for hydrogen ion exchange shown by lime-soda glasses can be

575

XIV, 9 QUANTITATIVE INORGANIC ANALYSIS

reduced, then other cations will become involved in the ion exchange process and
wc can see the possibility of an electrode responsive to metallic ions such as
sodium and potassium. The required cficct can be achieved by the introduction
of aluminium, and as shown in Table XI V. 1, this approach lias led to new glass
electrodes of great importance to the analyst.

In all ca.ses some sensitivity to hydrogen ions remains, and in any
potentioinetric determination with the.se modiikd gla.ss electrodes the hydrogen
ion concentration of the .solution must be reduced so as to be not more than 1 per
cent of the concentration of the ion being determined, and in a solution
containing more than one kind of alkali metal cation, some interference will be
encountered.

Tabic XIV, 1 Comiuisition of glasses for cation-seasiiiic glass' electrodes

Composition

For determination of

N'a.O C;iO SiO, 72 \

It ' l.Suhjet:l to alkaline error)

Lt,b 2s^..cs,or.., UjbaT, L.ijO, r.^sio. ojt

H' (Alkaline error icduecd)

LijO IS''', AI.Oj .SiO, 1)0',;

Lj’

N.ijO ii^. Ai,Oj ivS%.sid. 7r„

.N'a'.Ag"

NajO 27 “,. AI.Oj 5 SiO. OS

K‘

Tlie construction of these electrodes is exactly similar to that already described
for the pll-resporisive glass electrode. They must of course be used in
conjunction witii a reference electrode and for this purpose a siher-silver
chloride electrode is usually preferred. A ‘double junction’ reference electrode is
often used in which an inner tube eomaimng the siher- silver chloride electrode
immersed in potassium chloride solution is surrounded by an outer lube which
can be filled with any appropriate solution (potassium chloride, potassium,
sodium or ammonium nitrate, etc.), the choice being governed by the nature of
the solution under test. The inner tube is closed at its lower end by a porous
diaphragm, and the ba.se of the outer tube is closed by a ground-on glass cap;
electrical connection is achieved through tiie film of liquid trapped in the ground
joint. This arrangement makes it easy to change the junction liquid in the outer
tube, and it is claimed that this glass sea! type of junction gives a particularly
reproducible liquid junction potential and minimum ditTusion of junction
elecirolyte ituo the test solution; this last feature is important owing to the
possibility of interference from the ions thus introduced.

The electrode response to the ;iciiviiy of the appropriate cation is given by the
usual Ncrnsl equation:

£ = X-f-(«T//iF)log«„,.

and for a monovalent cation, since - log 0 ^,. = pM (cf. pH)
F = A:-0.0591 pM(al2.^X).

Such an electrode may however also show a response to certain other cations,
and when an intcrlcring cation B* *■ is present, then the expression for electrode
potential becomes:

^ 2.303 FT,

K-h — logu.M,

2303RT

576

POTENTIOMETRY XIV, 10/11

where «„=+ is the activity of the ion to be determined, is the activity of the
interfering ion, and is the Selectivity Coefficient of the electrode,

The value of this coefficient will- be determined by the nature of the interfering
species present: if the interfering ion is replaced by another interfering ion

then the Selectivity Coefficient will acquire a different value Hence
Selectivity Coefficients should always be denoted in the manner shown, which
indicates what particular interference is involved, and the activity of each species
must also be specified when quoting a value for the coefficient.

The selectivity coefficient can be evaluated by measuring the e.m.f. response of
the ion-sensitive electrode in solutions containing a constant activity of the
interfering ion B and varying activities of the principal ion M ; the smaller the
value of A:m,b the greater the preference of the electrode for the principal ion.

Values for the selectivity coefficients of a given electrode with respect to
common interferences are frequently quoted by the manufacturer, but it should
be noted that they are not always quoted in the recommended form (see Ref. 1 1).

XIV, 10. OTHER SOLID MEMBRANE ELECTRODES. The glass
membrane of the electrodes discussed above may be replaced by other materials
such as a single crystal or a solid ion exchange material ; it may be advantageous
to incorporate the ion exchange material into an inert carrier such as paraffin wax
or a suitable polymer.

Pungor (Ref. 2) developed an iodide-ion-sensitive electrode by incorporating
finely dispersed silver iodide into a silicone rubber monomer and then carrying
out polymerisation. A circular portion of the resultant silver iodide-impregnated
polymer was used to seal the lower end of a glass tube which was then partly filled
with potassium iodide solution (0.1 M), and then a silver wire was inserted to dip
into the potassium iodide solution. When the membrane end of the assembly is
inserted into a solution containing iodide ions, we have a situation exactly similar
to that encountered with glass membrane electrodes. The silver iodide particles in
the membrane set up an exchange equilibrium with the solutions on either side of
the membrane. Inside the electrode, the iodide ion concentration is fixed and a
stable situation results. Outside the electrode, the position of equilibrium will be
governed by the iodide ion concentration of the external solution, and a potential
will therefore be established across the membrane and this potential will vary
according to the iodide ion concentration of the test solution.

This original Pungor or heterogeneous membrane type of electrode has beeii
extended to give electrodes capable of measuring the concentration of Cl“, Br~,
CN , S^“, and many other anions, and electrodes suitable for measuring the
concentration of Q-, Br“, and 1“ can be obtained by using a membrane cast
from the appropriate pure silver halide; that is to say, the inert matrix is
dispensed with, and we are dealing with a solid state electrode. A particularly
useful application of this last technique is the single crystal lanthanum fluoride
electrode developed by Orion Research Inc, which can be used to measure the
concentration of fluoride ions in solution.

XIV, 11, LIQUID MEMBRANE ELECTRODES. Another type of selec-

ive ion electrode is based upon the use of liquid ion exchange materials, usually

577

XIV, 12 QUANTITATIVE INORGANIC ANALYSIS

consisting of tin ion exchange material dissolved in an organic solvent which is
not miscrblc with water to any great extent and thus obviating undue mixing of
the electrode material witli the solution to be analysed. Two dilTercnt types of
electrode arc used: (a) those in which the liquid e.xclianger contains the ion to
which the electrode is responsive, and (f>) those in which the liquid e.vchangcr is
electrically neutral and does not contain any ions.

Important electrodes of the first type are (i) the calcium-responsive electrode
based upon the calcium salt of didecyl hydrogen phospliatc dissolved in di-n-
octylphcnylphosphonate (Ref. 3), and (ii) the anion-responsive electrodes based
upon the methyl tri-ociunoyl-ammonium cation ( Ref. 4) : these are suitable, inter
iilia. for the determination of CIO^". SO 4 *", and many organic anions. An
example of the second kind of electrode is the Philips poliissium electrode in
which the ion exchange material is an antibiotic (vaiinomycin) dissolved in
diphenyl ether. Vaiinomycin forms an association complex with alkali metal ions
with the important feature that the selectivity coetficient for K* as compared
wiihNa *■ isaboutdOOO.undfor K.* ascompared with 11 is about IS 000 , so that
the electrode can be used to determine potassium in the presence of large
amounts of sodium, and in relatively strongly acid solutions.

In these liquid membrane electrodes, the solution of the ion c.xchange material
is placed in a tube closed by a porous diaphragm at its low er end, and the internal
silver-silver chloride electrode in potassium (or sodium) chloride solution is
placed in a narrow tube which is mounted inside the wider one.

In a recent development it has lieen shown (Ref. S) that if the active
components of a liquid membrane electrode (c.xchange medium plu.s solvent) are
added to a solution of polyvinyl chloride in teirahydrofuran, and the resulting
mixture allowed to stand for some days for the iciraliydrofuran to evaporate,
then a solid residue is left, from whieli a circle may be cm and cemented to the end
of a PVC tube. This arrangement then functions as a heterogeneous membrane
type of electrode, responding to the same ion(s) as the original liquid membrane
electrode.

XIV, 12. COMMERCIALLY AVAILABLE lO-VSENSIITVE ELEC-
TRODES. At the present time a number of ion selective electrodes are
available from laboratory supply houses and new ones arc frequently being
added: whilst not intended to be an exhaustive list, Table .XIV, 2 serves to
indicate the range of determinations for whicli electrodes are now available; see
Refs. 5, 6 , 7.

A range of gas-sensing elcctrode.s arc also available which can be used to
determine soluble gases such as hydrogen chloride, ammonia, sulphur dioxide,
and carbon dioxide. In the.se electrodes the gas stream i.s pa.ssed through a lube
containing a semi-perme;ible membrane separating it from a solution of carefully
selected pH. The soluble gas will pass through the membrane, dissolve in the
solution and thus ulleet tiic pH. The actual meas\iring electrode is a pH-
responsive glass electrode, and the measured change in pH can be related to the
concentration of the gas under investigation.

As already explained, some care must be exercised in using an ion-sensitive
electrode to ensure that interferences do not arise from other ions, and it is of
course also necessary to ensure that the ion which is to be measured has not
undergone complex formation with any of the reagents which have been added to
the solution; conversely, it may be possible to reduce the interference due to a

578

POTENTIOMETRY XIV, 13

Table XIV, 2. A selection of commercially
available ion sensitive electrodes.

Type ot
membraae

Ion

Lower limit of detection
(mol dm"’) ■

Glass

Na+

jq-6

K +

10-®

Liquid

K"-

io-«

Ca^’-

10-^

NO3-

10-=^

CIO4-

10-^

Solid

Ag-’-

10"”

lo-’

Cd^+

10-’

10*®

10"*

5x 10"*

Br-

5x10"*

5x10"®

CN'

10"*

SCN-

10"*

S’-

10'”

given ion by adding a reagent which will complex the interfering ion. As an
example, if it is required to measure the fluoride ion concentration of a solution
with a fluoride-responsive electrode, it is important to ascertain whether the
solution contains any aluminium which causes formation of the ion AlFg^ " ; the
fluoride ion which is thus bound will not affect the fluoride ion electrode. If the
solution does contain aluminium, it is treated with a complexing agent (e.g.,
cyclohexane-diamine-tetra-acetic acid) which complexes the aluminium and
releases the fluoride ion from the AlFg^' ion.

Instrumentation and measurement of cell e.m.f.

XIV, 13. POTENTIOMETERS. The most satisfactory method for the
measurement of the e.m.f. of a cell is that known as PoggendorfPs compensation
method, an outline of which is given below. The principle of the method is to
balance the unknown e.m.f. against a known e.m.T, which can easily be varied.
When these two e.m.f.s are exactly equal, no current will flow through a
galvanometer placed in the circuit : the galvanometer is therefore employed as a
null instrument. The essential details are shown in Fig. XIV, 5. ' >

A 2- or 4-volt accumulator furnishes the opposing e.m.f. ; this is connected in
series with a rheostat and with the terminals of a slide wire AB. The latter is a thin
wire of uniform cross-section, and is often termed the ‘potentiometer wire’. The
cell, the e.m.f. of which is to be determined, is connected to one end A of the slide
wire, and though a galvanometer G and akey to a sliding contact C, which can

579

XIV, 13 QUANTITATIVE INORGANIC ANALYSIS

be moved along .-IB. A .special double-
throw .switch S, may be provided to permit
the standard eel! to be placed in the circuit.
In connecting the accumulator and the cell
to the bridge, it is essential that the positive
poles should be connected to the same end
of the bridge wire; the unknown cell will
then send a current through the circuit in a
direction opposite to that furnished by the
accumulator.

If we assume that the potentiometer wire
has uniform cross-section and resistance,
then the fall of potential along the slide
wire will be uniform. TheditTerence of potential between A and any point C will
be proportional to the length AC, and will be equal to the fraction AC/^iBofihe
total fall of potential along the wire. If the standard cell is now' placed in circuit
and the position of C adjusted to say C so that when the switch 5'^ is depressed,
no current passes ilirougji the galvanometer G, then the e.m.f. of the cell is c-qual
to that of the accumulator multiplied by ACjAB. For much potentiometric
work, only chang<.'s of potential are required, so that for e.xample varialion.s of the
length /IC are all that arc required during a titration. In genenil, however, the
e.m.f. of the accumulator is not quite constant, and as indicated above a standard
cell is therefore emplo) ed to calibrate the slide wire. This is usually a Weston cell
which has an e.m.f. of 1.0183 volts at 20 C or

1. 0183- 0.0000406 (/- 20 )

at any other temperature / C. it should be noted however that there are often
small differences in the e.m.f. of Weston cells supplied by different makers, and it
is c-ssential to use the c-sact value quoted for the cell in use, or preferably the value
obtained by calibration of the cell in use against a Weston cell of known e.m.f.
which is reserved for calibration purposes.

If the standard cell is placed in circuit by means of the .switch .S’, and the point
of balance C on the bridge is determined, then the unknown e.m.f, may be
calculated from the expression ;

A C _ e.m.f. of unknow n cell

AC e.m.f. of standard cell

For approximate work, the slide wiie A B may consist of a simple meter bridge,
and the indicating instrument may be a milhammeter. It is preferable, however,
to employ a commercial type of potentiometer which utilises the more compact
spiral type of bridge wire.

1 he most convenient type of indicating instrument is the direct-vision type of
mirror galvanometer; the galvanometer, lamp, and scale are incorporated in a
blackened wooden (or plastic) box or compartment, and the ‘spot’ is clearly
visible in daylight; a switch on the front of the instrument allows a choice of
sensitivity. Alternatively, a solid state d.c. Null Detector (M. Tinsley and Co. or
Croydon Precision Instrument Co) may be used; these instruments have the
advantage of being less susceptible to vibration than the conventional mirror
galvanometer.

In commercial potentiometers, a rheostat is provided in series with the 2-volt

Fig. XIV, 5

580

POTENTIOMETRY XIV, 13

accumulator which can be adjusted so that the effective e.m.f. applied across the
potentiometer is such that scale readings are directly in volts (or millivolts). If the
bridge is divided into, say, 2000 equal parts, then the rheostat- may, be adjusted
with the standard cell, in circuit and with the sliding contact C at a position
corresponding to 1018.3 divisions so that no current flows through., the
galvanometer. The position of the sliding contact will then give the e.m.f. of any
unknown cell, directly in milli-volts. It is usual for the rheostat to contain both
coarse and fine adjustments : the fine adjustment may be used to compensate for
the slight variations of the accumulator during the measurements.

The Tinsley general-utility potentiometer (type 3387B) shown in Fig. XIV, 6 is
an excellent commercial potentiometer. Balancing is effected upon a main dial
having eighteen steps of 0. 1 volt and a calibrated circular slide wire range — 0.005
to +0.105 volt, which can be read to 0.0001 volt by estimation, the smallest
division being 0.0005 volt. The instrument has three range multipliers of x 1,
x0.1,and x 0.01, giving the following ranges of direct calibration; 1.9 volts to 1
millivolt, 0.19 volt to 100 microvolts, and 0.019 volt to 10 microvolts respectively.

Fig. XIV, 6

There is an independent standardising circuit, adjusted to 20 “C, so that
standardisation may be effected with a Weston cell independently of the dial
setting. The selector switch has three positions, one for the standardising circuit,
and two for external test circuits.

For work of the highest precision, one of the Tinsley 5590 series of Precision
Vernier Potentiometers may be employed; Model 5590C for example is a five-
dial potentiometer readable to seven decimal places, and with the smallest
subdivision equal to one microvolt. ■ ,

Whatever kind of potentiometer is .used, the working cell (the accumulator)
must be joined to the instrument which is then left for twenty to thirty min utes
before attempting any readings; during this period the resistance coils of the

581

XIV, 14 QUANTITATIVE INORGANIC ANALYSIS

insirumeiU may warm up sligluly and sufiicient time must be allowed for thermal
equilibrium to be established. The standard cell is then switched into circuit and
the balance point on the potentiometer slide vvire determined, or with an
instrument such as the Tinsley potentiometer which reads directly in volts, the
Selector switch is turned to the ‘Calibrate’ position, and the balancing rheostat is
adjusted until the galvanometer shows no deflection. Ai the commencement of
the standardising operation, the galvanometer sensitivity switch must be set to
the lowest .sensitivity, and then the sensitivity can be gradually increased as the
point of balance is approached. Finally, the standard cell is replaced by the test
cell (on the Tinsley potentiometer this is achieved by turning the Selector switch
to position ‘1’ or ‘2’, depending upon which pair of terminals the cell has been
joined to), and with the galvanometer at its lowest sensitivity, first the step-wise
control (steps of 0.1 V) is adjusted, and then the slide wire control until the
balance point is reached : once again as the balance point is approached, the
galvanometer sensiviiy is increased until ma.ximum sensitivity is achieved.

in much analytical work, the measurement of cell e.m.f. may be simplified by-
use of a solid slate millivollmeier (see pi I meters and plon nteters in the following
Sections); the e.m.f is thus obtained immediately front a single reading without
the need for the time-consuming balancing process needed with a conventional
potentiometer.

.XIV, 14, pH MIvTERS. In view of the high resistance of a glass electrode
(1-100 megohms) a simple potentiometer cannot be used to measure the e.m.f. of
a cell which includes a glass electrode, and in fact the c.irly development of the
gla.ss electrode was dependent upon advances in the design of thermionic valves
permitting the construction of ‘valve voltmeters*. Since these instruments were
designed with the requirement.s of the glass electrode in mind, and the glass
electrode was used to mea-sure the pH of .solutions, the instruments were referred
to as pH meters. L'arly pH meters were ciassilied as (u) direct reading, or (/>)
polentiomctric type meters. In meters of type (o) the e.m.f. of the cell containing
the glass electrode was impressed upon a high resistance and the current flowing
in tile re.sistance was then amplified and applied to a sensitive moving coil meter;
this was calibrated in millivolts so that the cell e.m.f was recorded directly, and
since in fact the quantity to be measured was pH, the scale was also calibrated in
pH units, a selector swilcli being provided to allow choice of scale reading. In
meters of type (/i) a poientionieiric circuit vv.is employed in conjunction with an
electronic amplifier and a miliiammclei as balance point detector. The
potentiometer was balanced against a standard cell contained within the
instrument, and then the e.m.f of the cell containing the glass electrode was
applied to the poteiuiomeler and balance achieved in the usual manner by
adjustment first ol a ‘coarse’ (stepwise) control, and then of a ‘fine’ (slide vvire)
control; these controls were calibrated in millivolts aiul also in pU units.

VVith the introduction ol solid state circuitry which hassimplilied the problem
ol nieasuring .small d.c. potentials in circuits of high impedance, the diavt
reading type ol pH meter i.s now standard; and in the most modem type of meter
a digital voltmeter is u.sed, scaled to read pH directly. Such instruments are
supplied by many makers; a typical c.\amplc is the Flextronic Instruments Ltd
Model 7060 pH meter shown in Fig. .XIV. 7.

As already explained, a glass electrode has an ‘asyninictry potential’ which
makes it impossible to relate a measured electrode potential directly to the pH of

582

POTENTIOMETRY XIV, 14

Fig. XIV, 7

the solution, and makes it necessary to calibrate the electrode. A pH meter
therefore always includes a control (‘Set Buffer’, ‘Standardise’ or ‘Calibrate^ so
that with the electrode assembly (glass plus reference electrode or a combination
electrode) placed in a buffer solution of known pH, the scale reading of the
instrument can be adjusted to the correct value.

The Nemst equation shows that the glass electrode potential for a given pH
value will be dependent upon the temperature of the solution. A pH meter
therefore includes a biasing control so that the scale of the meter can be adjusted
to correspond to the temperature of the solution under test. This may take the
form of a manual control, calibrated in °C, and which is set to the temperature of
the solution as determined with an ordinary mercury thermometer.. In some
instruments, arrangements are made for automatic temperature compensation
by inserting a temperature probe (a resistance thermometer) into the solution,
and the output from this is fed into the pH meter circuit.

Some instruments also include what is known as a ‘Slope control’. This is to
allow for the fact that in some cases, if a meter is calibrated at a certain pH (say
pH 4.00), then when the electrode assembly is placed in a new buffer solution of
different pH (say 9.20), the meter reading may not agree exactly with the known
pH of the solution. In this event, the slope control is adjusted so that the meter
reading in the second solution agrees with the known pH value. The meter is
again checked in the first buffer solution, and provided the scale reading is correct
(4.00), it is assumed that the meter will give accurate readings for all pH values
falling within the limits of the two buffer solutions. . .

Using a given glass electrode-reference electrode assembly, if we measure the
cell e.m.f. over a range of pH, all measurements being at the same temperature,
and if the readings are then repeated for a series of different temperatures,, then
on plotting the results as a senes of isothermal curves, we find that at some pH
value (pHj), the cell e.m.f. is independent of temperature; pH; was referred to by
Jackson (Ref. 9) as the ‘isopotential pH’. If the composition of the solution

583

XIV, 15 QUANTITATIVE INORGANIC ANALYSIS

surrounding the inner silver-silver chloride is altered, or if an entirely difl'erent
external reference electrode is used, then the value of pH, changes, and some pH
meters include an Tsopotential’ control which can be used to take account of
such changes in the electrode system.

Mode of operalion

Before use, it is obviously necessary to become familiar with the instruction
manual issued with the pH meter it is proposed to employ, but the general
procedure for making a pH measurement is similar for all instruments, and will
follow a pattern such as that detailed below.

1. Switch on and allow the instrument to warm up; the time for this will be
quite short if the circuit is of the solid-stale ty pc. Whilst this is taking place, make
certain that the requisite biilfer solutions for calibration of the meter are
available, and if necessary prepare any required .solutions: this is most
conveniently done by dissolving an appropriate ‘bulfer tablet' (these are
obtainable from many suppliers of pH meters and from laboratory supply
houses) in the specilied volume of distilled water.

2. If the instrument is equipped with a manual temperature control, take the
temperature of the solutions and set the control to this value; if aiuomatie
control is available, then place the temperature probe into some of the first
standard bulfer solution contained in a small beaker vvhicli has been previously
rinsed with a little of the solution.

3. Insert the electrode a.sscmbly into the same beaker, and if available, set the
selector. switch of the instrument to read pH,

4. Adjust the '.Set Bulfer' control until the meter reading agrees with the
known pH of the bulfer solution.

5. Remove tlie electrode assembly (and the ihermomclcr probe if used), rinse
in distilled water, and place into a small beaker containing a little of the second
buffer solution. Ifihc meter reading does not agree e.xactly with the known pH,
adju-st the 'Slope' control until the required reading is obtained.

6. Remove the electrode assembly, rinse in distilled water, place in the fir-st
buffer solution and confirm that the correct pH reading is shown on the meter: if
not, repeat the calibration procedure.

7. If the calibration is satisfactory, rinse the electrodes, etc., with distilled
water, and introduce into the teat solution contained in a small beaker. Read off
the pH of the solution.

S. Remove the electrodes, etc., rinse in di,-.lilled water, and leave standing in
distilled water.

XIV, 15. SELECTIVE ION .METERS, Direct reading meters suitable for
use with specific ion electrodes are available from a number of manufacturers:
they are sometimes referred to as ion activity meters. They are very similar in
construction to pH meters, and most can in fact be used as a pH meter, but by
virtue of the extended range of mcasuremenLs for which they must be used
(anions as well as cations, and divalent as well as monov alent ions), the circuitry
is necessarily more complex and scale expansion facilities arc included. typical
meter of this type is tlie Electronic Instruments' Ltd Model 7050 specific ion meter
sliown in Fig. XIV, 8.

As with a pH meter, the electrode appropriate to the measurement to be

584

POTENTIOMETRY XIV, 16

Fig. XIV, 8

undertaken must be calibrated in solutions of known concentration of the chosen
ion ; at least two reference solutions should be used, differing in concentration by
2-5 units of pM according to the particular determination to be made. The
general procedure for carrying out a determination with one of these instruments
is outlined in Section XIV, 17.

Direct potentiometry

The use of a pH meter or a specific ion meter to measure the concentration of
hydrogen ions or of some other ion in a solution is clearly an example of direct
potentiometry. In view of the discussion in the preceding Sections the procedure
involved will be evident, and two examples will suffice to illustrate the
experimental method.

XIV, 16. DETERMINATION OF pH. At this stage it should be pointed out
that the original definition of pH = — log Ch (due to Sorensen, 1909; this may be
written as pcH) is not exact, and cannot be determined exactly by electrometric
methods. It is realised that the activity rather than the concentration of an ion
determines the e.m.f. of a galvanic cell of the type commonly used to measure pH,
and hence pH may be defined as

pH=-logaH* . .

where 0 ^. is the activity of the hydrogen ion. This quantity as defined is also not
capable of precise measurement, since any cell of the type

Ha.Pt 1 H'*' (unknown) 1 Salt Bridge 1 Reference Electrode

585

XIV, 16 QUANTITATIVE INORGANIC ANALYSIS

used for the raeasurcrneiu inevitably involves a liquid junction potential of more
or IciS uncertain magnitude. Nevertheless the measurement of pi I by thee.m.f.
method gives values corresponding more closely to the activity than the
concentration of hydrogen ion. It can beshorvn that the pcH value is nearly equal
to -log 1.1 hence:

pH :..pdl+0.04

This equation is a useful practical formula for converting tables of pH based on
the Sorensen scale to an approximate activity basis, in line with the practical
ddiniiton of pH given below.

The modern definition of pH is an operational one and is ba.sed on the work of
siandardi-sation and the recommendations of the US National Bureau of
Standards (NBS). The .NBS definition and the British definition are consistent in
nearly all respects: the British defmiiion is coniined to one standard solution
whereas the NBS defmiiion c.vtcnds to a number of .standard solutions. The
dijfercmc 'xn pH between two solutions .S'(a standard) and A'(an unknown) at the
same temperature with the same reference electrode and with hydrogen
elecirode-s at the same hydrogen pressure is given by:

where Ex is the e.m.f. of the cell

H,,Pi 1 Solution ,\'! 3..').)/lvCl I Reference electrode,

£v is the e.m.f. of the cell

H,,Pl i Solutions i 3.5df KG 1 Reference electrode.

The pH dilference is liuis a pure number. The scale is anchored by defining the
nature of the standard solution and a,ssigning a pH value to it.

In the British standard. .V is a 0.0.3.!/ solution of pure potassium hydrogen
phthalale, the pH of which is d.OOOut 15 C. At any other temperature, i. between
0 and 55 'C the pH i.s given by:

In the NBS recommendations equal importance is given to ;i small number of
solutions for which pH values have been determined with great care over a range
of temperatures: one of these solutions is the potassium hydrogenphthalatc
solution on vvhieli the British standard is based, and the pH valued are almost
idenlicai. Data lor four ol the NBS standards now accepted by the lUP.AC are
collected in fable XIV, 3. The tartrate, pluhalatc, phosphate, and sodium
tetraborate solutions arc regarded as primary standards: at least two reference
solutions should be used for the standardisation of cells with the glass electrode.
Below pH 2 and above pH 12 the liquid-liquid Junction potential is suspect; for
this reason the tetroxalate solution and the calcium hydro,xidc solution are
designated as secondary standards, and the pH values obtained by their use may
be 0,02-0.04 unit lower than tho.se obtained with an instrument standardised
with the primary standards.

In passing it may be noted that the British Standard when applied to dilute

586

POTENTIOMETRY XIV, 16

solutions (<0.1M) at pH. between 2 and 12 conforms approximately to the
equation

pH= -log{cH»y,:i}±0.02

where y,.i is the mean activity coefficient which a typical 1: 1-electrolyte would
have in that solution.

Table XIV, 3. pH of NBS standards from 0 to 95 °C

Tem-

perature,

°c

Secondary

standard

Primary standards

Secondary

standard

|d'

T ^ 0

S Sec

0 a 14

^ va

^ w ^

14 73K

1 »— so

5 20

'TI ►r'

o' 0,14

^ X

X Cd

s s

0 0

.EC

X 0

H m

0 =3

0 Z

»r>

X ^

£3

3 cS

0 S'

1.67

—

4.00

6.98

9.46

13.43

1.67

—

4.00

6.95

9.40

13.21

1.67

—

4.00

6.92

9.33

13.00

1.67

—

4.00

6.90

9.28

12.81

1.68

—

4.00

6.88

9.23

12.63

1.68

3.56

4.01

6.86

9.18

12.45

1.69

3.55

4.02

6.85

9.14

12.30

1.69

3.55

4.02

6.84

9.10 '

12.14

1.70

3.55

4.03

6.84

9.07

11.99

1.70

3.55

4.05

6.83

9.04

11.84

1.71

3.55

4.06

6.83

9.01

11.70

1.72

3.55

4.08

6.83

8.99

11.58

1.72

3.56

4.09

6.84

8.96

11.45

1.74

3.58

4.13

6.85

8.92

1.77

3.61

4.16

6.86

8.89

1.80

3.65

4.21

6.88

8.85

1.81

3.67

4.23

6.89

8.83

The fifth NBS primary standard (KH2PO4, 0.00870m; Na2HP04, 0.0304m),
which only covers a very limited pH range (7.37-7.53), has not been included in
Table XIV, 3, and it may be noted that the five primary standard solutions, with
concentrations expressed on a molal basis (i.e., moles of solute per kilogram of
solution), have been adopted as international standards by the lUPAC. The pH
values for these international standards are quoted to the third decimal place
(Ref. 10), and the values given in Table XIV, 3 agree with these when rounded off
to the second place of decimals.

Details for the preparation of the solutions referred to in the table are as
follows (note that concentrations are expressed in molalities) : All reagents must be
of the highest purity, e.g., A.R. products. Freshly distilled water protected from
carbon dioxide during cooling, having a pH of 6.1-13, should be used, and is
essential for basic standards. De-ionised water is also suitable. Standard buffer
solutions may be stored in well-closed Pyrex or polythene bottles. If the
formation of mould or sediment is visible the solution must be discarded. .

0.05/n-Potassium tetroxalate. Dissolve 12.70 g of the dihydrate in water
and dilute to 1 kg. The salt KHC204,H2C204,2H20 must not be dried above

587

XIV, 16 QUANTITATIVE INORGANIC ANALYSIS

50“C. The solution is stable and the bulTer uipacity is relatively high.

Saturated potassium hydrogen tartrate .solution. The pH is insensitive to
changes of concentration and the temperature of saturation may vary from 22 to
28^C: the e.xcess of solid must be removed. The solution does not keep for more
than a few days unless a preservative (crystal of thymol) is added.

0.05m-Pot;is.siuin hydrogenphthalate. Dissolve 10.2! g of the solid (dried
below 1 30 °C) in water and dilute to 1 kg. The pH is not affected by atmospheric
carbon dio.xide : the buffer capacity is rather low. The solution should be replaced
after 5-6 weeks, or etirlier if mould-growth i.s apparent.

0,025ff/-Phospluite buJfcr. Dissolve 3.40 g of KfHPO^ and 3.55 g of
Na,HP 04 (dried for 2 hours at 1 10-130 ’C) in carbon dioxide-free water and
dilute to 1 kg. The solution is stable when protected from undue c.xposure to the
atmosphere.

O.Ol/ii Borax. Dissolve 3.8! g of sodium tetraborate NkuBjO-.lOH.O in
carbon dio.xide-free water and dilute to 1 kg. The solution should be protected
from c.xposure to atmospheric carbon dioxide, and replaced about a month after
preparation.

Saturated calcium hydroxide. solution. Shake a large e.xccs.s of finely divided
calcium hydroxide vigorously with water at 25 -C. lilter through a sintered glass
filler (poro.sity 3) and store in a polythene bottle. Entrance of carbon dioxideinto
the solution should be avoided. The solution should be replaced if a turbidity
develops. The solution i.s 0,0203.U at 25 T (pH 12.45), 0,02 11. \/ at 20 'C (pH
12.47), and 0.01'.)5.Uat 30 C(pH 12.44).

To measure the pH of a giten solution the normal procedure i.s to use a glass
electrode together with a saturated calomel reference electrode and to measure
Ihee.m.r. of the cell with a pH meter; the scale of the meter is calibrated to read
pH directly. The procedure for use of a pH meter has already been described in
Section XIV, 14; the instruction manual of the insirumenl av;iilablc for use
should however be consulted for details of minor variations in the controls
supplied. The glas.s electrode supplied witit the instrument shv)uld be standing in
distilled water: if for any reason it is neces.sary to m;ike use of a new electrode,
then this must he left soaking in distilled water for at least twelve hours before
measurements are attempted. Never handle the bulb of the electrode, and
remember that the assembly is nccess;inly somewhat fragile ;md treat it with
great care: in particular the electrode must always be supported within the
measuring vessel (special electrode stands arc usually supplied with pH meters)
and not allowed to stand on the base of the vessel.

Prepare the buffer solutions for calibration of the pH meter if these are not
already available; the pota.ssium hydrogenphthalate buffer (pH 4), and the
sodium tetraborate butler (pH 9,2) are the mo.st commonly used for calibration
purposes. The solutions cun be prepared in accordance with the details given
above, but the simplest procedure is to make use of the buffer tablets which can
be purchased.

Check whether the in.strumcnt supplied is ecpiippcd for automatic temperature
compensation, and, if so, that the temperature probe (resistance thermometer) is
available. If it is not so equipped, then the temperature of the solutions to be used
must be measured, and the appropriate setting made on the manual temperature
control of the insirumenl.

Proceed to mca.sure the pH of the given solution, following the steps outlined
in Section XIV, 14. On completion of the determination, remember to wash

588

POTENTIOMETRY XIV, 17

down the electrodes with distilled water, and to leave them standing in distilled
water. • . . . '

XIV, 17. DETERMINATION OF FLUORIDE. This determination in-
volves the use of a ion sensitive electrode (Sections XIV, 9-11) in conjunction
with an ion activity meter. The electrode must (as with the glass electrode used for
pH measurements), be calibrated, using solutions of the appropriate ion at
known concentrations. In view of the influence of ionic strength on activity
coefficients, it is important that the test and the standard solutions should be of
comparable ionic strength. When dealing with a test solution containing a single
electrolyte little difficulty will be encountered in arranging for the test and
standard solutions to be of similar ionic strength, but this may not be the case
when the test solution has arisen from an involved analytical procedure. In such a
case, an ionic strength adjuster bufler is added to both test and standard solutions
so as to achieve comparable values for the ionic strength in all solutions, the value
being governed by the ionic strength adjuster buffer rather than by the sample
itself; a number of electrolytes may be used in this fashion, due attention being
paid to ensure that errors do not arise , owing to complexation, or to poor
selectivity of the electrode in use with respect to the added ion in relation to the
ion whose concentration is to be determined. Whereas for pH measurements it
suffices to calibrate the glass electrode at two pH values, for ion-sensitive
electrodes it is advisable to plot a calibration curve by making measurements
with a number (usually five to six) of standard solutions of varying
concentration.

As an alternative to plotting a calibration curve, the method of standard
addition may be used. We first set up the appropriate ion-sensitive electrode
together with a suitable reference electrode in a known volume of the test
solution, and then measure the resultant e.m.f. (EJ. Applying the usual Nernst
equation we can say

£, = X+klogyiC,

where K is the electrode constant, k is theoretically 2.303RT/nF but in practice is
the experimentally determined slope of the E vj log C plot for the given electrode,
/, and C, are the activity coefficient and the concentration respectively of the ion
to be determined in the test solution. A known volume of a standard solution
(concentration CJ of the ion to be determined is added to the test solution, and
the new e.m.f. E^ is measured ; Q should be 50-100 times greater than the value
ofC,. For the new e.m.f. Fj we can write:

E, = K + /clog y^(F, C, + F, C 3 )/(F, + Ki)

where F, is the original volume of the test solution.

Provided that the first and second solutions are of similar ionic strength, the
activity coefficients will be the same in each solution, and the difference between
the two e.m.f. values can be expressed as

AE = (Ej -£,) = lclog(F,C. + FjC3)/C,(I^ + FJ

from whence C, = ^5

lO^Eik^l + VJV,)-VjV,

Hence provided the value of the slope constant k is known, the unknown
concentration C, can be calculated. , '

589

XIV, 17 QUANTITATIVE INORGANIC ANALYSIS

Procedure. Set op the ion activity meter (a digital pH/miilivoUmeler, c.g.,
Coming-EOL Model 109, used in the millivolt mode is equally satisfactory) in
accordance with the manual supplied with the instrument.

The electrodes required are a Buoride ion-sensitive electrode (e.g.. Coming
No. 476042) and a calomel reference electrode of the type supplied for use with
pH meters.

Prepare the following solutions.

Sodium Jluoride staiidurd.s. Using A.R. sodium fluoride and de-ionised
water, prepare a standard solution which is appro-vimatcly 0.054/ (2.1 gdm'^),
and of accurately known concentration (Solution A). Take lOcm^ of solution 4
and dilute to I diif’ in a graduated flask to obtain solution B which contains
approximately lOp.p.m. fluoride ion. 20cm^ of solution B further diluted
(graduated flask) to lOOcm^ gives a standard (solution Cl, containing
approximately 2 p.p.m. fluoride ion, and by diluting lOcnv* and 5cm^ portions
of solution i? to 100 ern^, we obtain .standards D and E, containing respectively I
and 0.5 p.p.m. F".

Total Ionic Slrciiittl) Adjustment Buffer (TISAB). Dissolve 57 cm’ A.R.
aa-tic acid, 5S g A.R. sodium chloride and 4 g cyclohexane diamino-tetra-aa'tic
acid (CDTA) in 500cm'’ of de-ionised water contained in a large beaker. Stand
the beaker inside a water bath fitted with a constant level device, and place a
rubber tube connected to the cold vvalcr tap inside the bath. Allow water to flow
slowly into the bath and discharge through the constant level: this will ensure
that in the subsct(ueni treatment the .solution in the beaker will remain at
constant temperature.

Insert into the beaker a calibrated glass electrode-calomel electrode assembly
which is joined to a pH meter, then with constant stirring and continuous
monitoring of the pH. add slowly sodium hydroxide solution (S.U). until the
solution acquires a pH of 5.0--5.5. Pour into a dm* graduated flask and make up
to mark w ith de-ionised water.

The resulting solution will exert a bufl'ering action in the region pH 5~6, the
CDTA will complex any polyvalent lon.s which may interact with fluoride, and by
viaue of its relatively high concentration the solution will furnish a medium of
high total ionic strength, thuv obviating the possibility of variation of c.ni.f.
owing to varying ionic strength of the test soliilion.s.

Pipette 25 cm* of solution B into a lOOcm* beaker mounted on a magnetic
stirrer and add an equal volume of TIS.AB from a pijwtte. Stir the solution to
ensure thorough mixing, stop the stirrer, insert the fluoride ion-calomel electrode
system and measure the c.m.f. The electrode rapidly comes tvv equilibrium, and a
stable e.m.f. reading is obtained immediately. Wash down the electrodes and
then insert into a second beakercontaining a solution prepared from 25 cm* each
of standard solution C and TISAB; 'read the c.m.f. Carry out further
determinations using the standards D and £.

Plot the observed e.m.f. values against the concentrations of the standard
solutions, using a semi-log graph p.iper which covers four cycles (i.c., spans four
decades on the log scale) ; use the log a\i.s for the concentrations which should be
in term.s of fluoride ion concentration. A straight line plot (calibration curve) will
be obtained. With incrca.sing dilution of the solutions there tends to be a
departure Irom the straight line: with the electrode combination and measuring
system referred to above, this becomes apparent when the fluoride ion
concentration is reduced to ca. 0.2 p.p.m.

590

POTENTIOMETRY XIV, 18

Now take 25 cm^ of the test solution, add 25 cm® TISAB and proceed to
measure the e.m.f. as above. Using the. calibration curve, the fluoride ion
concentration of the test solution may be deduced. The procedure described is
suitable for measuring the fluoride ion concentration of tap water in areas where
fluoridation of the supply is undertaken. - ■ .

Potentiometric titrations

XIV, 18. POTENTIOIVIETRIC TITRATIONS; CLASSICAL METHOD.
In the two previous sections dealing with direct potentiometry the procedure
involved measurement of the e.m.f. between two electrodes; an indicator
electrode, the potential of« which is a function of the concentration of the
ion to be determined, and a reference electrode of constant potential:
accurate determination of the e.m.f. is crucial. In potentiometric titrations
absolute potentials or potentials with respect to a standard half-cell are not
usually required, and measurements are made whilst the titration is in progress.
The equivalence point of the reaction will be revealed by a sudden change in
potential in the plot of e.m.f. readings against the volume of the titrating
solution; any method which will detect this abrupt change of potential may be
used. One electrode must maintain at a constant, but not necessarily known,
potential ; the other electrode must serve as an indicator of the changes in ion
concentration, and must respond rapidly. The solution must, of course, be stirred
during the titration. Simple arrangements for potentiometric titration are given
in Fig. XIV, 9, and in Fig. XIV, 10. In the former diagram, A is a reference

electrode (e.g., a saturated calomel half-cell), B is the indicator electrode, and C is
a mechanical stirrer (it may be replaced, with advantage, by a magnetic stirrer) ;
the solution to be titrated is contained in the beaker. When basic or other
solutions requiring the exclusion of atmospheric carbon dioxide or of air are
itrated it is advisable to use either a three- or four-necked flask or a tall lipless

591

XIV, 19 QUANTITATIVE INORGANIC ANALYSIS

beaker equipped as shown in Fig. XIV. 10. It i.s convenient to u.se as reference
electrode a compact calomel half'Cell as supplied with pH meters (see Fig. XIV,
3); nitrogen may be bubbled through the solution before and, if necessary, during
the titration.

The e.m.f. of the cell containing the initial solution is determined, and
relatively large increments (l-5cm^) of the tiirant .solution are added until the
equivalence point* is approached; the e.m.f. is determined after each addition.
The approach of the e.p. is indicated by a somewhat more rapid change of the
e.m.f. In the vicinity of the equivalence point, equal increments (e.g., 0.1 or
0.05cm^) should be added; tlie equal additions in the region of the e.p. are
particularly important when the equivalence point is to be detennined by the
analytical method described below. Sufticient lime should be allowed after
each adrlition for the indicator electrode to reach a reasonably constant potential
(to ai. ± 1-2 millivolts) before the ne.\t increment is introduced. Several points
should be obtained well beyond the e.p.

To measure the e.m.f. the electrode system must be connected to a
potentiometer: the simple .ilide wire system shown in Fig. XIV, 5 will often
suffice, but it will usually be more convenient to employ a commercial
potentiometer, as for c.xample tliat shown in Fig. XIV, 6. If the indicator
eiccirodc is a membrane electrode (e.g., a glass electrode), then a simple
potentiometer i.s unsuitable and either a pll meter or a selective ion meter must be
employed ; the meter readings may give directly tlie v arying pi 1 (or pM) values as
titration proceeds, or the meter may be used in the millivollmeter mode, so that
e.m.f. values are recorded. Used as a miliiv oltmeter, such meters can be used with
almost any electrode a.v-.embly to record the results of many different types of
polentiometric titrations, and in many ca-ses the instruments have provision for
connection to a recorder so that a comimious record of the titration results can be
obtained, i.c., a titration curve is produceil.

A number of commercial ‘poteniiomciric titration units' are available which
comprise an electrode system (frequently a selection of electrodes is offered)
together with a special stand providing supports for the electrodes, and one or
two burette holders. The ba.sc of the stand incorporates a magnetic stirrer, and in
some cases a liot plate, so that if necessary the solution to be titrated may be
heated. A separate unit embodies a poiemioinetcr and a compact galvanometer;
or if the potentiometer is designed for operation from the a.c, mains, the balance
point indicator may be a 'magic eye’ electronic indicator similar to that used in
mains-openilcd conductivity bridges (Section XV, 4).

XIV, 19. USE OF BI.METALLIC ELECI RODE SYSi'EMa A tungsten
electrode does not respond readily to changes in potential in certain
o.vidation-reduction systems (e.g., Cr,0,*'. Fe^') whereas a platinum elec-
trode does. Hence a platinum-tungsten couple can be used instead of the usual
combination of a platinum electrode and reference (e.g., calomel) electrode to
indicate the end point. Willi tlic Pt-W couple the tungsten appears to undergo
anodic oxidation and acts as a kind ofaitackable reference electrode’. The useof
such bimetallic systems is empirical, and the optimum conditions should be
established by trial. With the Pt~W pair the potential is small at first, remains at

• The abbreviation e.p. will be used for equivalence point.

592

— ^OTENTIOMETRY XIV, 20/21

this value until very near the equivalence point, when it usually increases slightly,
and then there is an abrupt change at the equivalence point. It is unsuitable for
the titration of very dilute solutions (< O.OOIM).

XIV, 20. POLARISED INDICATOR ELECTRODES. Some redox couples
(e.g., Cr207^“, Cr^"^; Mn04“, Mn^"^; and 8203^“, 8405^“) encountered in
titrimetric analysis are somewhat slow in establishing steady potentials at a
platinum electrode when the measurement is made in the ordinary way (i.e., with
zero current). To eliminate long waiting periods for the attainment of steady
potentials in cases of this kind polarised indicator electrodes, at which
electrolysis is forced to occur at a slow rate, may be used; polarised mono-
metallic and polarised bimetallic systems have been employed. The former
consists of a polarised metallic electrode and an unpolarised electrode (e.g., a
calomel electrode). A bimetallic system consists of two identical pure platinum
wires, one polarised anodically and the other cathodically with a polarising
current of the order of a few microamperes; these appear to behave as two
dissimilar metals, and their single electrode potentials respond in a different
manner. At all events a distinct change in behaviour is apparent at the
equivalence point, and if the potential difference between the electrodes is plotted
against the volume of reagent added the usual differential type of curve is
obtained. The potential difference developed at the end point may be of the order
of 100-200 millivolts. When one or both redox couples involved in the titration
reaction behave irreversibly a polarised electrode may show a considerably
different change in potential at the equivalence point than when the measurement
is made with an unpolarised electrode at zero current. If both couples behave
reversibly the potential change of the indicator electrode will be about the same
with or without electrolysis.

In potentiometric titrations with polarised electrodes the measured quantity is
the change in e.m.f. at constant current (compare Amperometric Titrations,
Chapter XVII, in which the change in current at constant applied e.m.f. is
measured). In some cases titration curves obtained with one or two polarised
electrodes exhibit larger changes in e.m.f. at the end point than curves obtained
with an unpolarised indicator electrode, e.g., in the titration of iodine with
thiosulphate ion using either two identical polarised platinum electrodes or a
single platinum electrode polarised cathodically. The use of polarised eleetrodes
can entail an error corresponding to the amount of electrolysis that occurs at the
electrode ; this is always present with a single polarised electrode, and also occurs
with two identical polarised electrodes whenever the titration couple behaves
irreversibly. The error can be reduced to negligible proportions by using small
electrodes and a small electrolysis current. The main value of polarised electrodes
IS for titrations involving irreversible couples where’ an unpolarised electrode is
often very slow in acquiring a constant potential whereas a polarised indicator
electrode reaches a steady potential quickly at constant current, and large
vanations of potential are observed at the end point.

XIV, 21. DIFFERENTIAL POTENTIOMETRIC TITRATION. As indicated
m Section XIV, 18, as the end point of a titration is approached the e.m.f. of the
more rapidly ; and as shown by curve (b) in Fig. XIV, 14 (Section
^ against V (the volume of titrant added) is a maximum

3 he end point. It is possible to measure directly A£'/A F as a function of V and

593

XIV, 22 QUANTITATIVE INORGANIC ANALYSIS

this procedure is referred to as dilfercntiai potcntioinclric titration. The desired
resuit is accomplished by placing two identical indicatorclccirodes (c.g. platinum
wires) in tliesoiution to be titrated, but one of tltesc (the ‘isolated' electrode) is in
a small portion of the liquid that is separated from the main body of tliesoiution,
and hence isolated from immediate reaction with the titrani. A
.simple device for this purpo.se is depicted in Fig. XIV. 1 1. A small
volume of the solution is withdrawn into a dropper provided with
platinum-wire electrodes us shown in the figure: the latter are
connected to a high-resistance galvanometer, serving as a
voltmeter. In practice, the titration is conducted by adding small
uniform increments of titrant. The potential ditfereiicc, cor-
responding to AK, is read after each addition, then the liquid
w ithin the bulb is c.xpellcd and tlie bulb is feltlled with a portion of
the main solution. Thus at each stage the isolated electrode is kept
one increment, A V, behind the electrode in the main solution, but
is allowed to catcli up before the ne.vi increment is added by
completely e.xixlling the liquid within the bulb. The value of A£ begins to increase
rapidly near the equivalence [loinl : the latter is indicated by the nia.xiinum value
of A. The main advantage of the dilfercntiai method is that it does not require a
reference electrode; it is slower and less convenient than the technique of titrating
to the equivalence point potential. Ditferential methods are not suited for
titrations where the electrodes in the solution reach equilibrium very slowly.

XIV, 22. AUTO.M.VnC POTUNTiOMETRlC I rrU.VriONS. As already
mentioned (Section XI\', 18). by joining a recorder to a mains-operated
potentiometer, it is possible to produce directly the titration curve relating to the
poteniiometric titration under intestigation. If tlie delivery of titrant from the
burette is linked to the movetnent of the recorder chart, then the process becomes
auiomaiic, and a number of firms market titration units which fulfil this function.
A typical example is the Mctrolim 'Potciiiiograph’ (.Model E5J6) shown in Fig.
XIV. 1 2, svhich includes the control unii/chari recorder linked to a motor-driven
piston burette, and to tlie electrode .isscmbiy of the titration vessel on the right-
hand side of the photograph; the pajrer feed of the recorder is coupled to the
motor drive of the piston burette, and the iven of the recorder follows the change
in e.in.t. of the electrode assembly. The insuuiucni will also provide a plot of
A£/AF.

Automation has also been e.vtended to stop delivery of the titrant when the
potential ol the indicator electrode attains the value corresponding to the
equivalence point of the particular titration involved; this feature is clearly of
great value when a number ol repetitive titrations have to be performed. It is
neccs.sary to carry out a preliminary e.vperiment to determine the equivalence
point potential ol the indicator electrode (or more precisely, the equivalence
point c.m.l, lor the indicator clcctrodc-standard electrode combination in use),
and to prevent over-shooting the end point provision must be made for reducing
the rate of addition ol titrant as the end-point is approached. VarioiLS control
units can be purchased (c.g., the Melrohrn End-Point Titrator E526) which carry
the requisite instrumentation. A potentiometer is included, and the linal
equivalence point e.in.t. to be attuiued is set on this. At the start of the titration,
the c.m.l. set up by the electrode assembly in the titration vessel vvil! be far
removed Irom the equivalence point value, and so a ‘difference potential' will

S'M

Fig. XIV, 12

XIV, 23 QUANTITATIVE INORGANIC ANALYSIS

exist between the electrodes and the pre-set reading olThe potentiometer; it is this
‘difference potential’ which controls the subsequent operations. The delivery of
titrant from the burette is controlled in some systems by a solenoid-controlled
valve, or with a piston burette, by the rale at which the motor drives the burette.
It is usually necessary to select another e.m.f., somewhat lower than the
equivalence point e.m.f., from which point the titrant will be added slowly, and
this value must ahso be set on tlie control unit. Tlie dials on the control unit are
frequently scaled in pi I units as well as in millivolts, so that for an acid-base
titration the controls may be set in term.s ol p5 1 rather than of e.m.l.

With the control unit .set up with the readings appropriate to the titration to be
carried out, a measured volume of the solution to be titrated is introduced into
the titration vessel and diluted to a suitable volume (usually 50- 1 OOenf*), and the
burette is charged with the titrant. The magnetic stirrer is set into operation, and
the titration started. Initially, with a large difference potential, titrant will be
added rapidly to the solution and this will continue until the e.m.f. is equal to the
prc-sclecled ‘change-over potential*. At this point the addition of titrant is
slowed down ; the piston burette will l>c driven more slowly, or in the case of the
solenoid-controlled valve burette, the valve is largely closed down, or, in some
systems, two burettes arc provided, one with a coarse jet, and the other with a fine
jet ; at the change-over point, the burette with the coarse jet is shut off completely
and the titration is completed through the fmc-jet burcllc. Addition of titrant will
finally cease when the difference potential between c)ecirode.s and potentiometer
disappears. M a further aid to avoid over-shooting the end -point, it is usually
recommended that the burette be provided with a drawn-out tip which is inserted
directly into thesolution to be titrated, and in such a position, that as the solution
is stirred, the liquid in the ncighbuuriiood of the burette tip is directed towards
the indicator electrode, which thus tends to cut olf delivery of lit rant a little on the
early side. With continued stirring and dispersal of the added titrant, if the end
point has in fact not been attained, a difference potential will reappear and the
burette will be actuated again : step-wise addition will thus continue until the true
end-point is reached. Aulo-iitrators are not suitable for use in cases where the
indicator electrode response i.s slow, or when the chemical reaction involved in
the titration is slow.

.XIV, 23. LOCATION OF END POI.NTS. When a titration curve has been
obtained (i.e., a plot of e.m.f. readings again.st volume of titrant added) cither by
manual plotting of the e.xperimental readings, or with suitable equipment,
plotted automatically during tlie course of the titration; it will in general be of the
same form as the neutralisation curve for an acid, i.e., an S-shaped curve as
shown in Fig. X, 2 (Section X, 12). The central portion of such a curve is' shown in
Fig. XIV, 13 and also in Fig, XIV, 14((i), and clearly tlie end-point will be located
on the steeply rising portion of the curve, and a will in fact occur at the point of
inneciion. Although when the curve shows a very clearly marked steep portion,
one can give an appro.xiniatc value of the end-point a.s being mid-way along the
steep part oJ the curve, it is geiieraily necessary to carry out some geometrical
construction in order to fix the end point exactly. Three procedures may be
adopted for this purpo.se;

(n) the method of bisection;

(h) tlic method of parallel tangents;

596

POTENTIOMETRY XIV, 23

(a) The method of bisection. This can be applied when the curve shows
reasonably good straight lines before and after the steep part of the curve. Each
of these straight lines is extended (the lower portion to the right, the upper portion
to the left), and then at suitable points vertical lines are erected, one to the right of
the steep part of the titration curve and one to the left. These vertical lines are then
bisected, and the midpoints joined; where the line joining the rriidpoints cuts the
titration curve is the end point of the titration.

(b) The method of parallel tangents. For this method a thin rigid plastic
sheet large enough to cover the titration curve is required, and on this is marked a
central horizontal line, together with a number of pairs of parallel lines drawn on
either side of the central line, i.e., line 1 (the first line above the central marker) will
be paired with the corresponding line 1' equidistant below the marker. The scale
of the titration curves normally encountered will determine the dimensions of the
markings on the plastic sheet, but up to about ten pairs of lines will usually be
adequate, and it will be found convenient for identification purposes to mark
different pairs of lines in different coloured inks. When the sheet has been
prepared, a thin slot is then cut along the central marker so that the point of a
pencil can be introduced.

The method is used when the portions of the curve on either side of the steep
portion shows a marked curvature, and the procedure is to lay the plastic sheet on
top of the titration curve in such a position that a given pair of the parallel lines
(easily identified if they are similarly coloured), are tangential to the upper and
lower parts of the titration curve. A pencil mark made through the central slot at
the point where it cuts the steep part of the titration curve then identifies the end-
point.

(c) The method of circle fitting. For this method also a thin rigid plastic
sheet will be needed on which is marked a series of circles of varying sizes ; the
circles may be drawn independently of they may be concentric, but in either case
a small hole must be drilled in the plastic at the centre of each circle (or at the
common centre) so that a pencil point may be J , '^d. The circles should

597

XIV, 23 QUANTITATIVE INORGANIC ANALYSIS

increase in diameter in steps of about 1
cm and a maximum diameter of about
30 cm usually suflkcs; if larger diameters
are found to be necessary then it is not
essential to draw the complete circle
thus obviating the need for an ex-
cessively large plastic sheet. The method
ofprocedure is indicated in Fig.XIV, 13;
the plastic sheet is laid on the titration
curve and its position altered until one
of the circle.s fits the lower bend in the
curve and the position of the centre of
the circle is marked on the titration
curve, 'rise sheet is then moved to the
upper bend of the curve and when a
circle which fits the bend is found, the
position of the centre is marked. The
marks indicating the two centres are
then joined by a straight line (XX in the
diagram), and where this lino cuts the
steep part of the titration curve is the end
point.

Unless the curve has been plotted
automatically, the accuracy of the re-
sults obtained by any of the above
procedures will be dependent upon the
skill with which the titration curve has
been drawn through the points plotted
on the graph from the experimental
observations. It is therefore usually

V. erx'

Fig. XIV, 14

considered preferable to employ analytical (or derivaiiie) methods of locating the
end point; the.se consist in plotting thelirst dcrivatn e curve (AH/A V against r),or
the second derivative curve (A'£/AI'*) against T). The first derivative curve gives
a maximum at the point of inileclion of the titration curve, i,c., at the end point,
whilst the second derviative (A^£,'AI'")iszcro at the point where the slopcof the
AH'/AF curve is maximum.

File procedure may be illustrated by the actual results' obtained for the
potentiomeiric titration of 25.0 cm* of ai. 0. l.\/-ainmonium ironftl) sulphate
with standard (0.1095A/)-cerium(lV) sulphate sojuiion using platinum and
saturated calomel electrodes;

Fe- " + Co'* ^ = Fe^ " + Ce-* ^

The results are collected in Table XIV. 4. as are also the calculated values for the
first derivative AH'/A F (iniliivolt/cni'*) and the second derivative d'H’/Af'^. It is
clear that lor locating the eiid-point. only the experimental figures in the vicinity
ol the equivalence point are required; all the observed results for the
potentiomeiric titration arc given for the sake of completeness. It is convenient,
and simplifies thcculcuiations, ifsinall equal volumes oftilrani are added in the
neighbourhood of the end-point, but this is not essential.

598

POTENTIOMETRY XIV, 23

TableXIV,4 Potentiometric titration of solution with 0.1095M-Ce‘‘^ solution,
using platinum and calomel electrodes , . .

solution

£(mV0

AEIAV

A"£/AF^

added, cm^ {V)

(mVIcm^)

1.00

373

10.5

5.00

415

4.6

10.00

438 -

4.2

15.00

459-

6.4

20.00

491

■ 12

21.00

503

22.00

523

22.50

543

22.60

550

22.70

557

100

22.80

565

100

200

22.90

575

150

500

23.00

590

300

1500

23.10

620

2400

21000

23.20

860

550

-18500

23.30

915

290

-2600

23.40

23.50

944

140

-1500

958

-840

24.00

986

40.5

-300

26.00

30.00

1067

14.5

-90

1125

, In Fig. XIV, 14, are presented : {a) the part of the experimental titration curve
in the vicinity of the equivalence point; (6) the first derivative curve, i.e., the slope
of the titration curve as a function of V (the equivalence point is indicated by the
maximum, which corresponds to the inflexion in the titration curve) ; and (c) the
second derivative curve, i.e., the slope of curve {b) as a function of V (the second
denvative becomes zero at the inflexion point and provides a. more exact
m^surement of the equivalence point).

Ihe optimum volume increment A V depends upon the magnitude of the slope
0 the titration curve at the equivalence point and this can easily be estimated
rom a preliminary titraton. In general, the greater the slope at the e.p., .the

599

XIV, 24 QUANTITATIVE INORGANIC ANALYSIS

smaller should AV be, but it should also be large enough so that the successive

values of A/T exhibit a significant difference.

When the titration curve is symmetrical about the equivalence point the end-
point defined by the maximum value of AEjAV is identical with the true
sloichiomeirical equivalence point. A symmetrical titration curve is obtained
when the indicator electrode is reversible and when in the titration reaction one
mol. or ion of the titrant reagent reacts with one mol. or ion of the substance
titrated. Asymmetrical titration curves result w'hen the number of molecules or
ions of the reagent and the substance titrated are unequal in the titration
reaction, e.g., in the reaction

5Fe*^" -f-MnO^' -i-.SH * — 5Fe^^ -f Mn*’ +411,0

In such reactions, even though the indicator electrode functions reversibly, the
maximum value of AEjAV will not occur exactly at the stoichiometric
equivalence point. The resulting titrntum error (dilTercncc betsveen end point and
equivalence point) can be computed or can be determined by experiment and a
correction applied. The titration error small when the potential change at the
equivalence point is large. With most of the reactions used in polentionielric
analysis, the titration error is usually small enough to be neglected. It is assumed
that suflicient time is allowed for the electrodes to reach equilibrium before a
reading is recorded.

As has been indicated, if suitable automatic titrators are used, then the
derivative curve may be plotted directly and there is no need to undertake the
calculations described above. Likewise if a differential titration is carried out
(Section .XIV, 21), then data is available which can be plotted directly to give the
first derivative curve.

When the potential of the indicator electrode at the equivalence point is
known, either from a previou-s experiment or from calculations, the end-poiiii
can be determined simply by adding the tiiram solution until this equivalence-
point potential is reached. This technique is analogous to ordinary titrations with
indicators and is very convenient and rapid. The potentiometer is set to this
potential, and the titrant solution is added {dropwi.se near the end point) until the
galvanometer shows no dellection, or reverses the direction of deficciion, when
the tapping key is closed momentarily. The accuracy of this technique will
depend upon the reproducibility of the equivalence point potential : it need only
be known approximately when AEVAITs large.

XIV, 24. SOME GENERAL CON'SIDEH.ATION’S. In thi.s and succeeding
Sections experimental details arc given for some typical potentiometric
titrations; with this information it should be possible to deduce the appropriate
procedure to be followed in other cases. The majority of potentiometric titrations
involve chemical reactions which can be classified a-s («) neutralisation reactions,
(h) o.xidution- reduction reactions, (c) precipitation reactions or (iJ) complc.xation
reactions, and for each of these different types of reaction, certain general
principles can be enunciated.

(a) NeutralLsation reaefions. The indicator electrode may be a hydrogen,
glass, or antimony electrode; a calomel electrode is generally employed as the
reference electrode.

The accuracy with which the end point can be found potenfiometrically
depends upon the magnitude of the change in c.rn.f. in the neighbourhood of the

600

POTENTIOMETRY XIV, 24

equivalence point, and this depends upon the concentration and the strength of
the acid and alkali (compare Sections X, 13-16). Satisfactory results are obtained
in all cases except; (a) those in which either the acid or the base is very weak
{K < 10“®) and the solutions are dilute, and (b) those in which both the acid and
the base are weak. In the latter case an accuracy of about 1 per cent may be
obtained in 0. 1 M solution.

The method may be used to titrate a mixture of acids which differ greatly in
their strengths, e.g., acetic and hydrochloric acids ; the first break in the titration
curve occurs when the stronger of the two acids is neutralised, and the second
when neutralisation is complete. For this method to be successful, the two acids
or bases should differ in strength by at least 10® to 1 .

(b) Oxidation-reduction reactions. The theory of oxidation-reduction
reactions is given in Sections II, 25. The determining factor is the ratio of the
concentrations of the oxidised and reduced forms of certain ion species. For the
reaction;

Oxidised form+n electrons Reduced form

the potential E acquired by the indicator electrode at 25 °C is given by ;

0.0591

log

[Ox]

[Red]

where is the standard potential of the system. The potential of the immersed
electrode is thus controlled by the ratio of these concentrations. During the
oxidation of a reducing agent or the reduction of an oxidising agent the ratio, and
therefore the potential, changes more rapidly in the vicinity of the end point of
the reaction. Thus titrations involving such reactions (e.g., iron(II) with
potassium permanganate or potassium dichromate or cerium(IV) sulphate) may
be followed potentiometrically and afford titration curves characterised by a
sudden change of potential at the equivalence point. The indicator electrode is
usually a bright platinum wire or foil, and the oxidising agent is generally placed
in the burette.

(c) Precipitation reactions. The theory of precipitation reactions is given
m Sections X, 29-30. The ion concentration at the equivalence point is
determined by the solubility product of the sparingly soluble material formed
during the titration. In the precipitation of an ion I from solution by the addition
of a suitable reagent, the concentration of I in the solution will clearly change
most rapidly in the region of the end-point. The potential of an indicator
electrode responsive to the concentration of / will undergo a like change, and
hence the change can be followed potentiometrically. Here one electrode may be
a saturated calomel or silver-silver chloride electrode, and the other must be an
electrode which will readily come into equilibrium with one of the ions of the
precipitate. For example, in the titration of silver ions with a halide, (chloride,
bromide, or iodide) this must be a silver electrode. It may consist of a silver wire,
or of a platinum wire or gauze plated with silver and sealed into a glass tube.

mce a halide is to be determined, the salt bridge must be a saturated solution of
potassium nitrate. Excellent results are obtained by titrating, for example, silver
ni rate solution with thiocyanate ions. Mechanical stirring is desirable to

accelerate the attainment of solubility equilibrium.

(d) Complexation reactions. In many cases of this type of titration,

orap ex formation results from the interaction of a sparingly soluble precipitate

601

XIV, 25 QUANTITATIVE INORGANIC ANALYSIS

witli an excess of reagent ; this occurs for example when we titrate a solution of
potassium cyanide with silver nitrate, where silver cyanide initially produced
dissolves in excess potassium cyanide to give the complex ion [Ag(CN),]' and
consequently only a very small concentration of silver ions. This situation
continues up to the point where at! the cyanide ion has been converted to the
complex ion. the increasing concentration of whicli also means a gradually
incretising concentration of free silver ions and consequently a gradual rise in the
potentiarof a silver electrode in the solution. At the end point, there is a marked
rise in potential which enables the end point to be determined, but if the addition
of silver nitrate is continued past this point, the e.m.f. changes only very
gradually and silver cyanide is precipitated. F-inaily a second rapid change in
potential is observed at the point where all thecyatiide ion hxs been precipitated
as silver cyanide. For this particular titration a silver electrode is the obvious
indicator electrode, and as relerenec electrode either a mcrcury-mercury(i)
sulphate electrode, or a calomel electrode which is isolated from the solution to
be titrated by means of a pota.ssium nitrate or potassium sulphate salt bridge.

For complexation titrations involving the use of FDTA. an indicator electrode
can be set up by using a mercury electrode in the presence of mercurj{UFEDTA
complex (see Section XIV, 29).

XIV, 25. SO.ME EXPEHIME.NTAL DETA112> FOR POTENTIO.METRIC
TITRATIONS. A lew simple experiments will be brictly de.scribcd, the
performance of which will enable the reader to obtain experience of the
technique. Experiment 1 will require the use of a pH meter (or specific ion meter)
which should be employed in the millivolt mode, and it is suggested that
experiment 2 be carried out u.sing a .simple potentiometer to measure the e.ra.f.
According to the avaiiafailiiy ofapparatus, tiic other experiments may be carried
out using a commercial potentiomctric titration apparatus (manual measure-
ment of e.m.f.), and witli a commeicial appar.itus which plots the titration curve
automatically. In this way a wide range of experience vvill be acquired, but of
course if need be, all the experiments can be carried out using a simple
potentiometer, apart from the experiment involving use of a glass electrode for
which a pH meter is esscmi;d.

Expcrinwiit I. .NeutralLsation reaetioas. Prepare solutions of acetic acid
and of sodium hydroxide, each appro.ximatcly 0.1.1/ and set up a pH meter as
described in Section .XIV, 14,

The following general iuslruciions arc applicable to most potentiomctric
titrations and are given in detail here to avoid subsequent repetition.

Ui) Fit up the apparatus shown in Fig. XJ V. 9 with tlie electrode as-scmbly (or
combination electrode) supplied with the pH meter supported inside the beaker.
The beaker has a capacity of about 400cm-* and contains SOenf* of the solution
to be titrated (the acetic acid).

(/;) Select a burette, and by means of a piece of polythene tubing attach to the
jet a piece of glas.s capillary tubing aboui h*~10cm in kngth. Charge the burette
with the sodium hydroxide solution taking care to remove all air bubbles from
ihecapillary extension, and then clamp the burette so that the end of the capillary
is imnienscd in the solution to be titrated. This procedure ensures that all
additions recorded on the burette have in fact been added to the solution, and no
drops have been left adhering to the tip of the burette; a factor which can be of
some significance lor e.m.f. readings made near the end point of the titration.

602

POTENTIOMETRY XIV, 25

(c) Stir the solution in the beaker gently. Read the potential difference
between the electrodes with the aid of the meter. Record the reading and also the
volume of alkali in the burette.

(d) Add 2-3 cm^ of solution from the burette, stir for about 30 seconds, and,
after waiting for a further half minute, measure the e.m.f. of the cell.

(e) Repeat the addition of 1-cm^ portions of the base, stirring and measuring
the e.m.f. after each addition until a point is reached within about 1 cm^ of the
expected end-point. Henceforth, add the solution in portions of 0. 1 cm^ or less,
and record the potentiometer readings after each addition. Continue the
additions until the equivalence point has been passed by 0.5-1 .0 cm^.

(/) Plot potentials as ordinates and volumes of reagent added as abscissae ;
draw a smooth curve through the points. The equivalence point is the volume
corresponding to the steepest portion of the curve. In some cases the curve is
practically vertical, one drop of solution causing a change of 100-200 millivolts
in the e.m.f. of the cell ; in other cases the slope is more gradual.

(g) Locate the end-point of the titration by plotting AE/AV for small
increments of the titrant in the vicinity of the equivalence point ( F = 0. 1 cm^ or
0.05 cm^) against V. There is a maximum in the plot at the end point (compare
Fig.XIV,14(Z))).

(h) Plot the second derivative curve, A^EIAV^, against V: the second
derivative becomes zero at the end-point (compare Fig. XIV, 14(c)). This
method, although laborious, gives the most exact evaluation of the end-point.

Other suggested experiments include titration of 0.05M-Na2C03 with O.IM-
HCl, and ofO. lAf-boric acid in the presence of 4 g of mannitol with 0. 1 Af-NaOH.

Experiment 2. Oxidation-reduction reaction. Experience in this kind of
titration may be obtained by determining the iron(II) content of a solution by
titration with a standard potassium dichromate solution.

Prepare 250cm^ of O.lA^-potassium dichromate solution (using the dry A.R.
solid) and an equal volume of ca. 0AM ammonium iron(II) sulphate solution;
the latter must contain sufficient dilute sulphuric acid to produce a clear solution,
and the exact weight of A.R. ammonium iron(II) sulphate employed should be
noted. Place 25 cm^ of the ammonium iron(II) sulphate solution in the beaker,
add 25 cm^ of ca. 2.5Af-sulphuric acid and 50 cm^ of water. Charge the burette
with the O.liV-potassium dichromate solution, and add a capillary extension
tube. Use a bright platinum electrode as indicator electrode and an S.C.E.
reference electrode. Set the stirrer in motion. Proceed with the titration as
directed in Experiment 1. After each addition of the dichromate solution
measure the e.m.f. of the cell. Determine the end-point: (i) from the
potential-volume curve and (ii) by the differential method. Calculate the
molarity of the ammonium iron(II) sulphate solution, and compare this with
the value computed from the actual weight of solid employed in preparing
the solution.

Repeat the experiment using another 25 cm^ of the ammonium iron(II)
sulphate solution but with a pair of polarised platinum electrodes. Set up two
small platinum plate electrodes (0.5 cm square) in the titration beaker and
remove the two electrodes previously in use. Connect the platinum plates to a
polarising circuit consisting of a 50 volt dry battery joined to a 20 megohm
resistor so that a minute current will flow between the electrodes when they are
P aced in solution. Also join the electrodes to the circuit used for measuring the
ce e.m.f.; a simple potentiometer which is excellent for the first part of the

603

XIV, 25 QUANTITATIVE INORGANIC ANALYSIS

experiment cannot be used with the polarised electrodes, and the most
satisfactory procedure is to use the millivolt scale of a pH meter. Some
commercial pptentiometric titration units make provision for titration with
polarised electrodes. The end-point of the titration Ls indicated by the large jump
in e.m.f.

The experiment may also be repealed using a platinum (indicator) electrode
and a tungsten wire reference electrode. If the tungsten electrode has been left idle
for more than a few days, the surface must be cleaned by dipping into just molten
sodium nitrite (CAIiEl). The salt sliould be only just at the melting point or the
tungsten will be rapidly attacked ; it should remain in the melt for a few seconds
only and is then thoroughly rvashed with distilled water.

Experiment 3. Precipitation reactions. The indicator electrode must be
reversible to one or the other of the ions which is being precipitated. Thus in the
titration of a potassium iodide solution with standard silver nitrate .solution, the
electrode must be cither a silver electrode or a platinum electrode in tlie presence
of a little iodine (Isest introduced by adding a little of a Ireshly prepared alcoholic
solution of iodine), i.e,, an iodine electrode (reversible to P). The exercise
recommended is the stand.irdisation of silver nitrate solution with pure sodium
chloride.

Prepare an approximately 0. ld/-silver nitrate solution. Place 0.1 169 g of dry
A.R. sodium chloride in the beaker, add lOOcm^ of water, and stir until
dissolved. Use a silver-wire eleeitode (or a sihcr-plated platinum wire), and a
silver-silver chloride or a saturated calomel referenee eleetrode separated from
the solution by a potassium nitrate- agar bridge (sec below). Titrate the sodium
cliloride solution with the silver nitrate solution following the genera) procedure
described in Experiment I ; it is important to have efficient stirring and to wait
long enough aHcr each adiiiiion ul' titrant for tlie e.m.f. to become steady.
Continue the titration to 5cm^ beyond the end-point. Determine the end-point
and thence the molarity of the silver nitrate solution.

The salt bridge vviiich is required in tins c.xperiment is prepared from a piece of
narrow glass tubing which is first bent at right angles giv'ing a limb long enough to
reach to near the bottom of tbx titration vessel, flic tube is then given a second
right angle Ixnd in such a position that the horizontal limb will extend from the
titration vcs.scl to a suitable position in which a small beaker can be supported;
the two vertical limbs of the bridge should be of equal length. Clean the tube
thoroughly and then clamp with the two vertical limbs extending upwards.
Dissolve 3 g of agar in lOtlenv* of hot (almost boiling) distilled water, and then
add dOg of A.R. potassium nitrate. .As soon as the .salt has dissolved, allow to
cool for a lew minutes, and then carefully pour the hot liquid into the inverted
bridge tube so that it is filled eompletely and with no air bubbles emrained in the
liquid; a drawn-out thistle funnel will be found useful for this operation. Allow
tlie tube to cool completely still in the inverted position, and when cold it may be
found that at the ends of the tube the gel h;is conttMcted somewhat, so that when
the tube is placed in a liquid, an air bubble is trapped at the bottom of the tube; if
thi.s has happened, the extreme ends of the lube should be carefully cut off.

For this particular titration, a three- or four-necked Hat bottom flask is
conveniently used as titration vcs.sel; the .salt bridge can then be inserted into one
of the necks ol the llask and held in position by mean.s of a cork. The free end of
the bridge is allowed to dip into a .small beaker containing potassium nitrate
solution (33/), and the side arm of the reference electrode is then inserted into the

604

POTENTIOMETRY XIV, 26

beaker. When not in use, the salt bridge should be stored with the two ends
immersed in potassium nitrate solution contained in two test tubes. A potassium
chloride-agar bridge is obtained by replacing the potassium nitrate by 40 g of
A.R. potassium chloride.

An interesting extension of the above experiment is the titration of a mixture of
halides (chloride/iodide) with silver nitrate solution. Prepare a solution ( 1 00 cm^)
containing both potassium chloride and potassium iodide ; weigh each substance
accurately and arrange for the solution to be about 0.025M with respect to each
salt. A silver nitrate solution of known concentration (about 0.05A/) will also be
required.

Pipette 10 cm^ of the halide solution into the titration vessel and dilute to
about 100 cm^ with distilled water. Insert a silver electrode, an agar-potassium
nitrate salt bridge and complete the cell with a saturated calomel electrode. Fit a
10 cm^ micro burette with a capillary tube extension, and fill with the silver
nitrate solution. Add 1 cm^ of the silver nitrate solution to the contents of the
titration vessel and read the cell e.m.f. after allowing adequate time for the value
to become stable ; complete the titration in accordance with the details previously
given, but remember that there will be two end points, one in the neighbourhood
of 5cm^ of silver nitrate (I“), and the other in the neighbourhood of lOcm^
(Cl-)..

A comment on the polarity of the electrodes of the silver-calomel electrode cell
may be helpful at this point. With the respective values of electrode potentials
(calomel 0.245 V, Ef^ 0.799 V) one would normally expect the silver electrode to
be the positive electrode of the cell, but at the start of the above titration, the
concentration of silver ions in solution is so minute that the log term in the Nernst
equation in fact has a large negative value, and the potential of the silver
electrode actually becomes smaller than that of the calomel electrode. With
continued addition of silver nitrate, the concentration of silver ions in solution
gradually rises and the potential of the silver electrode increases, and at a point
which occurs near the first end-point of the titration it becomes equal to, and
subsequently greater than, the potential of the calomel electrode. When this
point is reached, it is necessary to reverse the connections of the leads from the
cell to the potentiometer, and in order to plot a satisfactory titration curve the
subsequent readings must be regarded as negative: conversely, of course,
the initial readings may be regarded as negative, and those after the change
over point as positive.

XTV, 26, DETERMINATION OF COPPER. Prepare a solution of the
sample, containing about 0. 1 g copper and no interfering elements, by any of the
usual methods ; any large excess of nitric acid and all traces of nitrous acid must
be removed. Boil the solution to expel most of the acid, add about 0.5 g urea (to
destroy the nitrous acid) and boil again. Treat the cooled solution with
concentrated ammonia solution dropwise until the deep-blue cuprammonium
compound is formed, and then add a further two drops. Decompose the
cupramnionium complex with glacial acetic acid and add 0.2 cm^ in excess. Too
great a dilution of the final solution should be avoided, otherwise the reaction
be^een the copper(II) acetate and the potassium iodide may not be complete.

Place the prepared copper acetate solution in the beaker and add 1 0 cm^ of 20
per cent potassium iodide solution. Set the stirrer in motion and add distilled
water, if necessary, until the platinum plate electrode is fully immersed. Use a

605

XIV, 27/28 QUANTITATIVE INORGANIC ANALYSIS

saturated calomel reference electrode, and carry out the normal potenlioraetric
titration procedure using a standard sodium thiosulphate solution as titrant.

XIV, 27. DE’l'ERMlN.VriON OF CHROMIUM. The chromium in the
substance is converted into chromate or dichromatc by any of the usual methods.
A platinum indicator electrode and a saturated calomel electrode arc used. Place
a known volume of thcdichromatesolution in the titration beaker, add lOcra^of
10 percent sulphuric acid or hydrochloric acid per 1 00 cm^ of the final volume of
the solution and also 2..5 cm-' of 10 i>crcent phosphoric acid. Insert the electrodes,
stir, and after adding Icm-' of a standard ammonium iron(II) sulphate solution,
the e.m.f. is mea.sured. Continue to add the iron solution, reading the e.m.f. after
each addition, then plot the titration curve and determine the end point.

XIV, 28. DETERMIN'.VITON OF MANGANESE. The method is based
upon the titration of manganese(II) ions with pennanganate in neutral
pyrophosphate solution :

4Mn-" fMn 04 ' -i- Sir +1.511, P,0,-" = 5Mn(lI,P,0,)j^ - +4H,0

The manganese{lll) pyrophosphate comple.v has an intense reddish-violet
colour, consequently the titration must be performed poieniiometrically. A
bright platinum indicator electrode and a saturated calomel reference electrode
may be used. The change in potential at the equivalence point at a pM between 6
and 7 is large (about .100 millivolts); the potential of the platinum electrode
becomes constant rapidly after each addition of the potassium permanganate
solution, thus permitting direct titration to almost the equivalence point and
reducing the time required for a determination to less than 10 minutes. With
relatively pure manganese solutions, a sodium pyrophosphate concentration of
0.2-0.31/, a pi 1 between 6 and 7, the equivalence point potential is +0.47 + 0.02
volt IX the saturated calomel electrode. At a pH above 8 the pyrophosphate
complex is unstable and the method cannot be used.

The method is at least as accurate as the bisinuthate procedure (Section X, 99)
and is even less subject to interferences. L,;irgc amounts of chloride, eobalt(Il),
and chromium(lU) do not inierferc; iron(lU), nickel, molybdenumlVl),
tungsien(Vi), and uraniumi VI) are innocuous; nitrate, sulphate, and perchlorate
ions arc harmle.ss. Large quantities of magnesium, cadmium, and aluminium
yield precipitates which may co-prccipitatc manganese and should therefore be
absent. Vanadium causes diliicultie-s only when the amount is equal to or larger
than the amount ol manganese; wium it is present originally in the +4 state, it is
o.xidised slowly in the titration if the + 5 state along with the manganese. Small
amounts of vanadium (up to '-.bout one-fifth of the amount of manganese) cause
little error. Hie interference of hirgc amounts of vanadium(V) can be
circumvented by performing the titration at a pH of .3-.3.5. Oxides of nitrogen
interlcre because of their reaction with potassium permanganate: hence when
nitric ;icid is used to dissolve the sample, the resulting solution must be boiled
thoroughly and a small amount of urea or sulphamic acid must l>e added to the
add solution to remove the last traces of o.xides of nitrogen before introducing
the sodium pyrophosphate solution.

For initial practice in the method determine the manganese content of
anhydrous A.R. mangancsc(ll) sulphate. Heat A.R. nianganesc(II) sulphate
crystals to 280 ‘C, allow to cool, grind to a fine powder, reheat at 2S0 "'C for 30

606

POTENTIOMETRY XIV, 29

minutes, and allow to cool in a desiccator. Weigh accurately about 2.2 g of the
anhydrous manganese(II) sulphate, dissolve it in water and make up to 250 cm^ in
a graduated flask.

Prepare a 0.02M solution of potassium permanganate and standardise it
against A.R. arsenic(III) oxide.

Prepare 5M-sodium hydroxide solution using the A.R. solid: test a 10-cm^
sample for reducing agents by adding a drop of the permanganate solution; no
green coloration should develop.

Prepare also a saturated solution of the purest available sodium
pyrophosphate (do not heat above 25 °C, otherwise appreciable hydrolysis may
occur); 12 g of the hydrated solid Na 4 P 2 O 7 , 10 H 2 O will dissolve in 100-150 cm^
of water according to the purity of the compound. It is essential to employ freshly
made sodium pyrophosphate solution in the determination.

Place 150 cm^ of the sodium pyrophosphate solution in a 250-400-cm^ beaker,
adjust the pH to 6-7 by the addition of concentrated sulphuric acid from a 1-cm^
graduated pipette (use the appropriate indicator test-paper or a pH meter). Add
25 cm^ of the manganese(II) sulphate solution and adjust the pH again to 6-7 by
the addition of 5jVf-sodium hydroxide solution. Introduce a bright platinum
electrode into the solution, and connect the latter through a saturated potassium
chloride bridge to a saturated calomel electrode; complete the assembly for
potentiometric titrations as in Fig. XIV, 9. Stir the mixture, add the potassium
permanganate solution in 2-cm^ portions at first, reduce this to 0. 1-cm^ portions
in the vicinity of the end-point ; determine the potential after each addition. Plot
the e.m.f. values (ordinates) against the volume of potassium permanganate
solution added (abscissa), and determine the equivalence point. From your
curve read off the potential at the equivalence point; this should be +0.47 volt.
Calculate the percentage of Mn in the sample.

1 cm^ 0.02M-KMnO4 s 0.00439 g Mn

Further practice may be obtained by determining manganese in a manganese
ore and in a steel.

Pyrolusite. Dissolve 1 .5-2 g, accurately weighed, pyrolusite in a mixture of
25 cm^ of l:i hydrochloric acid and 6cm^ concentrated sulphuric acid, and
dilute to 250 cm^. Filtration is unnecessary. Titrate an aliquot part containing
80-100 mg manganese: add 200 cm^ freshly prepared, saturated sodium
pyrophosphate solution, adjust the pH to a value between 6 and 7, and perform
the potentiometric titration as described above.

Steel. Dissolve 5 g, accurately weighed, of a steel in 1 : 1 nitric acid with the
aid of the minimum volume of hydrochloric acid in a Kjeldahl flask. Boil the
solution down to a small volume with excess of concentrated nitric acid to re-
oxidise any vanadium present reduced by the hydrochloric acid: this step is
unnecessary if vanadium is known to be absent. Dilute, boil to remove gaseous
products, allow to cool, add I g of urea or sulphamic acid, and dilute to
250 cm . Titrate 50-cm^ portions as above.

^’OTENTIOMETRIC EDTA TITRATIONS WITH THE

RCURY ELECTRODE. Discussion. The indicator electrode employed is
a mercury | mercury(II)-EDTA-complex electrode. A mercury electrode in
^ solution containing metal ions M"'*' (to be titrated) and a small
auded quantity of a mercury(II)-EDTA complex HgY^- (EDTA = Na 2 H 2 Y)

607

XIV, 29 QUANTITATIVE INORGANIC ANALYSIS

exhibits a potential corresponding to the liall'-ccll:

It can be shown that the potential at equilibrium is given by:

r-’O JiJ. [HgY^"} a:
n«‘MU+ 2F (MY‘"-'’^)' A',

[HgY-"} ^

where A'^y and A'„jv arc the stability (or formation) constants of the
nictal-EDTA and mercury-EDTA complexes respectively. The first two terms
on the right-hand .side of thi.s equation arc c.ssenlially constant during a
potentionietric titration, especially in the region of the end point, hence the
measured potential of the electrode becomes a linear function of pM. The
mercury I mercurydD-EDTA complex electrode will, for convenience, be
subsequently described as a mercury indicator electrode; it is clearly a pM
indicator electrode.

The potential of the mercury indicator electrode depends upon the total
mercury concentration in the solution. In practice it is found that the addition of
1 drop of a O.OOl-O.Ol .\/ solution of the mercury-EDTA complex HgY*" is
sufficient to establish a rea.sonably con.stam value for the mercury content so that
trace additions of thi.s metal do not seriously alter the shape of the titration curve.
In complexomctric titrations of metal ions with the mercury electrode, the
experimental conditions, sucli as pH, the kind and nature of the bulTer solution,
must be carefully controlled. The Iniffer should be prc.sent in an amount sufticieni
to prevent pH changes during tiie titration: a large excess of bulfer should be
avoided, as this may decrease the extent of the potential break al the end-point.
Halide ions mu.st not lie present in appreciable concentration.s because they may
interfere with the electrode reaction, especially for titrations performed under
acid condition.s (c.g., chloride interferes at a pH !c.s.s than about 6.5). At a pH
lower than 2. the mercury( 11) -EDT.A complex di.ssociates to such an extent that a
poorly defined titration curve results. At a pH above about 11, o.xygcn reacts
w'ith mercury, leading to a distorted titration curve; this may be often avoided by
bubbling nitrogen through the solution before and during the titration. Direct
and also back-titration procedures have been used for the determination of
numerous metal ions and a selcetion of ^he^c is given below.

Apparatus, .\ferairy c/ccfrui/c. The electrode, together with the essential
dimensions, is shown in Fig, ,\1V, 15; it is ea.sily constructed from Pyrex tubing.

, The platinum wire dipping into the mercury may be

welded to a copper wire; but it is preferable to use a
platinum wire siilficieiuly long to proirudeal the topof
the electrode tube. The mercury must be pure and
clean; in case of doubt, the mercury should be washed
with dilute nitric acid and then thoroughly rinsed with
distilled water. The electrode is filled with mercury
from above, ami it is allowed to pass into the annular
space through the hole until the outside compartment
is almost filled. It is most important that no mercury is
spilled into the titration vessel during the titration.
After each titration the electrode is repeatedly washed

with distilled water.

Fig. XIV, 15

Alterjiatively, an amalgamated gold electrode may

608

POTENTIOMETRY XIV, 29

be employed. This may be prepared by dipping a commercial gold plate electrode
for about a minute into pure mercury; after rinsing with water, it is ready for
use. The electrode can be used only a few times, and then must be re-
amalgamated.

Titration assembly. The electrode system consists of a mercury electrode and a
saturated calomel (or, in some cases, a mercury-mercury(I) sulphate) reference
electrode, both supported in a 250-cm^ Pyrex beaker. Provision is made for
magnetic stirring and the potential is followed by means of a precision slide wire
potentiometer and a sensitive ‘spot’ galvanometer or better with a
potentiograph.

Reagents required. Standard EDTA solution, 0.05M. See Section X, 50.

Mercury-EDTA solution. Mix small equal volumes of 0.05M-mercury(II)
nitrate (prepared- from the A.R. solid) and O.OSM-EDTA ; neutralise the
liberated acid by the addition of a few drops of 3M-ammonia solution. (In acid
solution an insoluble precipitate, probably HgH 2 Y, forms after a few days).
Dilute 10.0 cm^ of this solution to lOOcm^ with distilled water. The resulting ca.
0.0025M-mercury-EDTA solution is used for most titrations.

Ammonia buffer solution. Mix 20 g ammonium nitrate and 35 cm^
concentrated ammonia solution, and make up to lOOcm^ with distilled water.
Dilute 80 cm^ to 1 dm^ with distilled water. The pH is about 10.1.

Acetate buffer solution. Mix equal volumes of 0.5Af-sodium acetate solution
and 0.5Af-acetic acid solution. The resulting solution has a pH of about 4.7.

Triethanolamine buffer solution. Prepare a ca. 0.5M aqueous solution of
triethanolamine and add 2.5Af-nitric acid until the pH is 8.5 (use a pH meter).

Procedure. The general procedure is as follows. Place 25.0 cm^ of the
metal-ion solution (approximately 0.05M) in a 250-cm^ Pyrex beaker, add
25 cm^ of the appropriate buffer solution and 1 drop of 0.0025ylT-
mercury-EDTA solution (1 drop of 0.025jl/-mercury-EDTA solution for
calcium and magnesium). Use the titration assembly described above. Stir
magnetically. Titrate potentiometrically with standard 0.05Af-EDTA solution
added from a burette supported over the beaker. Reduce the volume of EDTA
solution added to 0. 1 cm^ or less as soon as the potential begins to rise ; wait for a
steady potential to be established after each addition. Soon after the end-point
the change of potential with each addition of EDTA becomes smaller and only a
few large additions need be made. Care must be taken that mercury is not spilled
into the solution either during the insertion of the mercury electrode or during
the titration.

Plot the titration curve (potential in millivolts V 5 . S.C.E. against volume of
standard EDTA solution) and evaluate the end-point. In general, results
accurate to better than 0.1 per cent are obtained. Brief notes on determinations
with various metal-ion solutions follow.

Calcium. 25.0 cm^ calcium-ion solution-f 25 cm^ ammonia or tri-
ethanolamine buffer.

Magnesium. 25.0 cm^ magnesium-ion solution -f 25 cm^ ammonia buffer.

Nickel. 25.0 cm^ nickel-ion solution + 25 cm^ ammonia buffer.

Cobalt. 20.0 cm^ cobalt-ion solution -f- 25 cm^ ammonia buffer. *

Copper. 25.0 cm^ of copper(II)-ion solution +25 cm^ of acetate buffer.

A larger excess of ammonia buffer is required to ensure the formation of the cobalt-ammine.

609

XIV, 30/31 QUANTITATIVE INORGANIC ANALYSIS

Mercurj'. 25.0 cm^ inercury(II)-ion solution + 25 ciir^ acetate bufier.

Zinc. 25.0cm-* zinc-ion solution + 25 cin^ acetate buiFer.

Bisiuutli. 25.0 cm^ bi.sinuth-ion soluiioiw- solid hexainine to pH about
4.6: the precipitate of basic bi.snnith salt di.ssolves as the EDTA solution is added
but the titration is slow.

Lead. 25.0 cm* lead-ion .solution -(-solid hexaminc to pH about 4.6.

Thorium, (i) 25.0cm* thorium-ion solution -(-90 cm* O.OOl.U-nitric acid,
3.U-ammonia solution added until pH about 3.2; mercury -mercury(I) sulphate
reference electrode.

(ii) 25.0cm* thorium-ion .solution -(-solid he.xamine to pH about 4.6;

40.0 cm* 0.05,l/-EDT.-\ added, and back-tiirate e.xcess with standard lead nitrate
solution.

Chroiuiiim. 25.0cm* chromium(lIl)-ion solution (0.02.U, prepared by
dilution of stock solution) + 50.0cm* 0.02.U-EDT.\ + 50cm* acetate bulTer,
boiled for 10 minutes, solution ci)olcd, pH adjusted to 4.6 with he.xamine, I drop
of mercury-EDTA solution added, and then back-titrated with
standard zinc-ion solution.

Aluininiuni. 25.0 cm* aluminium-ion solution, acidilied with a few drops
of 2.54/-nitric acid (to pH 1-2). boiled for 1 minute, 50.0cm* 0.05.H-EDTA
added to hot -solution, .solution cooled, 50cm* acetate bulfer and 1 drop of
0.00253/-mercury-. EDTA added, c.xces.s of EDTA back-titrated with standard
zinc-ion solution.

XIV, 30. DETERMINATION OF IRON! HI) MTTH EDTA. f)ijcm.sw/!.
Thi.s is an example of a potenliomctric titration involving two diirereni oxidation
states of the same metal. Thus in the titration of iron(Ill) with EDT.A, the
potential measured is that of the I'e(lU)-Fe(ll) couple (compare Section X,3I.
and Variamine Ulue in Section 28). The titration isconducted at pH nv. 3 using
a bright platinum indicator electrode: the iron(ll) remains uncomple.xed with
EDTA during the titration and a large potential change accompanies the abrupt
change of iron(lll) concentration in the vicinity of the end-point. Metals which
form complexes with EDTA interfere.

ProiCiiurc. For practice in this deicmiinalion, prepare a
at. 0.U5.l/-atnmonium iron( HU sulphate solution by dissolving about 6.0 g of the
.A.R. salt, accurately weighed, m 25()cm* of water in a graduated flask. Dilute

25.0 cm* of this solution to 100 cm* with distilled water, and add dilute ammonia
solution dropwi.se until the pH is about 3.0. Titr.ite potentiometrically with
standard 0.05.l/'-EDTA (Section X, 50): use a small platinum foil as the indicator
electrode and a saturated calomel half-cell a.s the reference elc'ctrodc.

XIV, 31. STANDARDlS.VriON OF PO I ASSIUM PERMANGANATE
SOLUTION WITH POT'A.SSU’M IODIDE. Discitssioii. Potassium per-
manganate solution may be standardised very accurately by potenliometric
titration with A.R. potas.siuin iodide. The latter is only slightly hygroscopic; it
may be dried, if nece.s.sary, by healing at 200 'C. The titration apparatus consists
ol a 250-cm* Fyrex beaker (or, better, a four-necked flask), a bright platinum foil
electrode (1.2 x 1.0cm)and a mercury-mercuryfl) sulphate reference electrode, •
a saturated calomel electrode is also satisfactory.

It is claimed that the results by the potassium iodide method are reproducible
to ±0.01 percent and agree with thearsenic(lll)o.xideprocedure(ScctionX, 92)

610

POTENTIOMETRY XIV, 32

to within 0.03 per cent; the standardisation with arsenic(III) oxide may also be
performed potentiometrically.

Procedure. Weigh out accurately about 0.35 g dry A.R. potassium iodide,
dissolve it in 50 cm^ distilled water in the titration vessel, and add 1.0 cm^
concentrated sulphuric acid from a dropper pipette; dilute further to about
100 cm^. Stir magnetically and also bubble a slow stream of nitrogen through the
solution. Assemble the titration apparatus described above. Titrate poten-
tiometrically with The potassium permanganate solution {ca. 0.1iV):to be
standardised.

XIV, 32. DETERMINATION OF NICKEL AND OF COBALT BY
COMPLEXATION WITH CYANIDE. Discussion. The concentration of
the cyanide solution is first determined by potentiometric titration with standard
silver nitrate solution using a silver indicator electrode and a
mercury-mercury(I) sulphate reference electrode. Two points of inflection
(indicated by a rapid fall in potential) will be found in the titration curve
corresponding to the reactions ;

Ag++2CN-=[Ag(CN)2]-

[Ag(CN)2]-+Ag-^ = 2AgCN(i)

The silver-ion concentration at the second point of inflection is almost exactly
twice that at the first inflection point ; the latter is employed for the calculation of
the cyanide concentration.

For the determination of nickel, a nickel(II) salt in ammoniacal solution
is treated with excess of potassium cyanide solution, and the excess of the latter
is titrated potentiometrically with standard silver nitrate solution:

[Ni(NH 3 )J^+ +4CN- = [Ni(CN)J^- + 4 NH 3

For the determination of cobalt, a cobalt(ll) salt in almost neutral solution is
treated with excess of potassium cyanide solution, and the excess of the latter is
titrated potentiometrically with standard silver nitrate solution. The cobalt(II)
cyanide complex is pentacovalent [Co(CN) 5 ]^“, and this fact must be borne in
mind when calculating the cobalt-ion concentration. Somewhat less satisfactory
results are obtained in slightly ammoniacal solution.

Reagents. Silver nitrate solution, ca. O.IM. Weigh out accurately about
8.5 g A.R. silver nitrate, dissolve it in water, and dilute to 500 cm^ in a graduated
flask.

Potassium cyanide solution, ca. 0.7M. Weigh out about 6.5 g A.R. potassium
cyanide (CAUTION !) and dissolve it in 1 dm^ of water in a graduated flask.

Nickel-ion solution, ca. 0.05M. Weigh out accurately about 2.9 g pure nickel
pellets, dissolve in the minimum volume of concentrated nitric acid, boil gently to
remove nitrous fumes, cool, and dilute to 1 dm^ with distilled water in a
graduated flask.

Cobalt-ion solution, ca. O.OSM. Weigh out about 14.0 g A;R. cobalt(II)
sulphate and dissolve it in 1 dm^ of water in . a graduated flask. Determine the
exact cobalt content by titration with standard 0.05M-EDTA using Xylenol
Orange as indicator. ... ...

Apparatus. The titration vessel may consist of a 250-cm^ Pyrex beaker;
provision is made for magnetic stirring. A silver rod (3 mm diameter) is used as an
indicator electrode, and a mercury-mercury(I) sulphate half-cell as a- reference

611

XIV, 33 QUANTITATIVE INORGANIC ANALYSIS

electrode; the latter may be replaced by a saturated calomel electrode connected
to the titration vessel by means of an agar-potassium sulphate bridge.

The pota.ssiunt cyanide solution is conveniently measured out tvith a pipette
attached to a ‘Punipctt’; if tiie iatter i.s not available, a burette may be used.
Attention is directed to the hifihly poisonous character of the potassium
cyanide solution; the hands should always be thoroughly washed immediately after
handling the reagent.

Procedure. Standardisation of potassium cyanide .solution. Place 25.0 cm^
of the potassium cyanide solution in the titration vcs.sel and dilute to lOOcm^
with distilled water. Stir magnetically. Titrate potentiometrically in the usual
manner with standard 0.13/-silver nitrate .solution. Plot the titration curve
(potential in millivolt.s against volume of silver nitrate solution) and evaluate the
end-point (at the first sharp change in potential) using the first or second
dilferential curve.

Calculate the concentration of the potassium cyanide solution.

Determination of nickel. Place lU.OOcnf* of the nickel-ion solution in the
titration vessel, dilute to about lOOciif* with distilled water, and add 63/-
ammonia solution until the pH is about 10. Add 50.0cm^ of the potassium
cyanide solution, and titrate the excess of potassium cyanide potentiometrically
with standard O.l.U-siher nitrate. Evaluate the end-point of the latter titration.
Calculate the volume of potassium cyanide solution which has reacted with the
nickel, and thence the nickel content of the .solution (Ni- ‘ h 3CN').

Determination of cobalt. Place lO.OOcm^ of the cobalt-ion solution in the
titration ves,scl, and dilute to UXIcm-* with distilled water. .Add 50.0 cni^ of
the potassium cyanide solution, and titrate potentiometrically the e.xcess
of pota.ssium cyanide with standard 0.1,\/-silver nitrate. Deduce the volume of
pota.ssium cyanide solution winch has reacted with the cobalt and then calculate
thecobalt content of thesolution (Co’ ' a 5CN").

X1V,33. DEn'K.AlINATTON OK FLUOlUDK BY A .NULL-POINT
METHOD. Di.scuision. The method is based upon the comple.xing by
lluoride of one of the o.xidation states of a redox couple: the Ce(lV)-Ce(ni)
couple is generally used. The potential measured i.s that between two half-cells,
each initially containing the same volume of a Ce(lV)-Ce(llI) .solution. The
sample solution is added to one half-cell and an equal volume of water to the
other. The magnitude of the e.m.f. of the cell gives some measure of the fluoride
concentration. Standard fluoride solution is added to the second cell until the
e.m.f. ol the cell is zero. During the .utdition of the titrant, water is added at the
same rate to the first half-cell so that concentrations, etc,, remain similar in each
half-cell. For the analysis of O.J5.1/-fluoride solutions, the optimum working
conditions are with a fluoride; ccriumdV): cerium(IIl) ratio of 2;i:l. The
iippro.ximate fluoride concentration of the sample solution must be known, and
this can be evaluated very approximately by a preliminary titration.

Sulphate and nitrate do not interfere, nor does chloride in amount about equal
to the fluoride concentnition. Bromide interferes owing to the reduction of
cerium(lV) to cerium(UI); this can be overcome by using an iron(lll)-iron(ll)
rcdo.x couple. Acetate gives slightly high results; o.xalate, molybdate, and
phosphate interfere and must be absent. Alkali metals and ammonium have little
influence on the titrations. Other cations should be removed by passage through
an ion exchange column or by precipitation: the latter method is applicable in

612

POTENTIOMETRY XIV, 34

those cases (e.g., silver chloride from silver fluoride) where co-precipitation of
fluoride is negligible.

Reagents. Cerium(IV)-ceriiim(III) solution. Dissolve 6.35g pure am-
monium ceriumllV) sulphate dihydrate in 200 cm^ water and 14 cm^ 18M-
sulphuric acid. Then add 2.8 g pure cerium(III) sulphate, stir until dissolved, and
dilute the resulting solution to 1 dm^ with distilled water. The solution is O.OIM
in cerium(IV) and in cerium(III) and 0.25M in sulphuric acid.

Standard sodium fluoride solution, 0.05M. Prepare from dry A.R. sodium
fluoride.

Apparatus. The two half-cells consist of two 250-cm^ Pyrex or polythene
beakers, standing on two small but similar magnetic stirrers, and connected by an
agar-potassium chloride bridge. A clean platinum wire electrode is supported in
each beaker; the electrodes are connected to a precision slide-wire potentiometer.

Procedure. For practice in this determination, the fluoride content of A.R.
potassium fluoride may be determined : prepare a ca. 0.05 Af solution from an
accurately weighed amount of the dry solid, and regard this as the sample
solution.

Into each of the two dry and clean 250-cm^ beakers, place 50.0 cm^ of the
cerium(IV)-cerium(III) solution. Connect the two half-cells and check that the
e.m.f. of the cell is zero. Pipette, say, 20.0 cm^ of the sample solution into one
half-cell, and add the same volume of distilled water to the other. The magnitude
of the e.m.f. at this stage gives an approximate measure of the fluoride
concentration of the sample. Add the standard fluoride solution portion-wise
from a burette to the second half-cell. During the titration add distilled water
from another burette to the first half-cell at the same rate as the titrant is
introduced. Stir the contents of each beaker magnetically at about the same rate
throughout the titration. After each addition, measure the potential difference
between the electrodes. When the end-point is approached (i.e., when the e.m.f. is.
about 10 millivolts, and for about 1 cm^ on either side of the, equivalence point),
run in the titrant in 0.2-cm^ portions.

Plot the values of the e.m.f. against the volume of the standard sodium fluoride
solution, and read from the graph the exact volume required to produce zero
e.m.f. Calculate the fluoride content of the sample of potassium fluoride.

XIV, 34. References

1. G. Mattock (.1961). pH Measurement and Titration. London ; Heywood and Co. Ltd.

2. E. Pungor, J. Havas and K. Toth (1965). Z. Chem., 5, 9.

3. J. W. Ross (1967). Science., 156, 1378.

4. J. Ruzicka and J. C. Tjell (1970). Anal. Chim. Acta., 49, 346 ; 51, 1 .

5. A. K. Covington (1970-3). Electrochemistry. Vols. I/III. Specialist Periodical
Reports. London; The Chemical Society.

6. N. K. Lakshminarayanaiah (1972). Electrochemistry. Vol. 11. Specialist Periodical
Report. London; The Chemical Society.

7. G. J. Moody and J. D. R. Thomas (1973). Selected Annual Reviews of the Analytical
Sciences. Vol. 3. London; Society for Analytical Chemistry.

8. G. J. Moody and J. D. R. Thomas (1974). Chemistry and Industry., 644.

9. J. Jackson (1948). Chemistry and Industry, 1.

10. International Union of Pure and Applied Chemistry (1969). Manual of Symbols and
Terminology for Physicochemical Quantities and Units. London ; Butterworths

11. G. J. Moody and J. D. R. Thomas (1971). Talanta, 18, 1251.

613

XIV, 35 QUANTITATIVE INORGANIC ANALYSIS

XIV, 35. Selected bibliography

1. I, M. KoIlholTand N. H. Funnan {1931). Potetuiumetric Titrations, 2nd cdn. New
York; John Wiley.

2. H. H. Willard, L. L, Morriu and J. A. Dean 1 1974). Instrumoniul Methods of Analysis,
5th cdn. New York: Van Noiirand.

3. A. L. Bcilby (1970). Modern Clasdcs in Analytical Chemistry. Washington DC;
American Chetniea) Society.

4. L Meites (1963). llandhook if Analytical Cheinhiry. New York; .McGraw-Hill.

5. T. S. West (1972). .Amilyiical Chemistry. MTP Scries 1, Vols, 12/13. London;
Buticrworths.

6. H. A. Strobel (197.3). Chemical Instrumentation —A Systematic Approach (o
Instrumental Analysis. 2nd cdn. Reading, Mass.; Addlion-Weslcy Pub. Co.

7. H. F. Walton and J. Reyes (1973). .Modern Chemical Analysis and liistnmteniation.
New York ; Marcel Dckkcr Inc,

8. D. J. G. Res and G. J. Janz (1961 ). Reference Etectro.Ies. London: Academic Press,

9. R. G. Bates (1973). Determinaiion of pi/. 2nd cdn. New York; Wiley.

10. G. Eisenman (1967). Glass Eiectrode.s fur llydroyen and Other Cations. New York;
Marcel Dckkcr Inc.

11. R. A. Durst (1969). Ion Selectisc Electrodes. SpcciA Publieaiion No. 314. Washington
DC; National Bureau oi'Standafds.

12. A. J. Bard (from 1966). Electroanalyiical Chemistry. A Series of Advances. Various
volumes. New York; .Marcel Dckkcr l«c.

13. R. P. Buck (1974). 'Ion .Selective Electrodes, Putentiometry, Potcniiomctric
Titrations'. Review article. Analytical Clu mistry, 46, 2;5R.

14. J. W. Robinson (1970). Undergraduate Instrumental Analysis. New York; Marcel
Dckkcr Inc.

15. K. Caminann (1973). D.is .Irheiten mit lancn.ielektiten Elckiroden. Anleitungen fur
dicchemische Laboratonunispr.ais. Vol. XI II. Berlin; Springer V’erlag.

16. T. Anf.ili and D. Jagner (1973). 'Cotnpiit.ition of Intrinsic End-point Errors in
Titrations with Ion Selective Elecir odes’. , \nai. Chem., 45. 2412.

17. lUPAC (1975). Rectiinmendeahms far SomencLiture of hui'Selecihe Electrodes.
Appendices on Provtskmai Nomencl.rturc. Symbols, Units and Stiindards, No. 43.
O.sford.lUPAC Secretariat,

18. G. J. Moody and j. D. R. Thoma'-(197i). .Se.Wtiic /on Senstthc Electrodes. Watford;
Merrow Publishing Co, Ltd.

614

COWDUCTOMETRIC TITRATIONS

CHAPTER XV

XV, I. GENERAL CONSIDERATIONS. Ohm’s law states that the current
I (amperes) flowing in a conductor is directly proportional to the applied
electromotive force E (volts) and inversely proportional to the resistance R
(ohms) of the conductor:

7 = E/R

(compare Section XII, 1). The reciprocal of the resistance is termed the
conductance (G): this is measured in reciprocal ohms (ohm~^), for which the
name siemens (S) has been proposed (Ref. 1). The resistance of a sample of
homogeneous material, length /, and cross-sectional area o, is given by:

R = p.lla

where p is a characteristic property of the material termed the resistivity
(formerly called specific resistance). In SI units, / and a will be measured
respectively in metres and square metres, so that p refers to a metre cube of the
material, and

p = R.a/l

is measured in ohm metres. Hitherto, resistivity measurements have been made in
terms of a centimetre cube of substance, giving p the units ohm cm. The
reciprocal of resistivity is the conductivity, k (formerly specific conductance),
which in SI units is the conductance of a one metre cube of substance and has the
units ohm“^m“^ (or Sm“^), but if p is measured in ohm cm, then k will be
measured in ohm" ^ cm" ^ (or S cm" ^). Virtually all the data at present recorded
in the literature is expressed in terms of k measured in Scm"^ units, and these
values will therefore be adopted in this book. Furthermore, as pointed out in
Section XII, 1, most of the existing data is expressed in terms of the ‘international
ohm’ and not the SI ‘absolute’ unit introduced in 1 948.

The conductivity of an electrolytic solution at any temperature depends only
on the ions present, and their concentration. When a solution of an electrolyte is
diluted, the conductivity will decrease, since fewer ions are present per cm^ of
solution to carry the current. If all the solution be placed between two electrodes
1 cm apart and large enough to contain the whole of the solution, the
conductance will increase as the solution is diluted. This is due largely to a
decrease in inter-ionic effects for strong electrolytes and to an increase in the
degree of dissociation for weak electrolytes. . , •

.615

XV, 2 QUANTITATIVE INORGANIC ANALYSIS

The molar conducthity (A) of an electrolyte is defined as the conductivity due
to one mole and is given by :

A = 10()0k/C = 1000 V.

where C is the concentration of the solution in moles per diif*, and V is the
dilution in dm^ (i.c., the number of dm^ containing one mole). Clearly, since k
has the dimensions Scm"'. the units of A are Sciirmol'’, or in SI units,
Sm’mol"'.

For strong electrolytes the molar conductivity increases as the dilution is
increased, but it appears to approach a limiting value known as the molar
conductivity at infinite dilution A, ; this quantity is written as Aq when
concentration, rather tlian dilution, is considered. The quantity A^ can be
determined by extrapolation fur dilute solutions of strong electrolytes. For weak
electrolytes the e.vlrapolation method cannot be used for the detemtination of A(,
but it may be computed from ilie molar conductivities at infinite dilution of the
respective ions, use being made of the 'law of independent migration of ions’. At
infinite dilution the ions arc independent of each other, and each contributes its
part to the total conducliviiy. thus;

Aq ~ Ay (cat) + An (an)

wliere Ay (cat) and Ao (an) are the ionic molar conduclivitii;.s at infinite dilution of
the cation and anion respectively. The values for the limiting ionic molar
conductivities for some ions m w.iter at 25 ' C arc collected in Table .W, 1.

Table XV, I I.iniiliugiouic molar conductivities at 25 C

Cation

A„(cat)

Aalun

A»(ao)

11*

W.H

01 {-

198.3

Na*

50.1

55 4

K.*

n.i

7t,,t

Lt'

38.7

llr

78.1

Nil,'

73.5

76 «S

Ag'

61.0

NO,

71 S

Tl'

74.7

CIO/

1)4,6

KV*

50,5

CIO.

67 4

].Sr* '

i').5

lirO.

55.7

tlU-'

63.6

10,

AH 5

IMg' '
IZn-’

53 I

lo.-

54.6

52 H

ncu,'

44.5

69.5

ICO/’-

fi9 }

(Cu-*

53.6

ISO/-

so.o

iNi-’ •

ii’O.^ ■

iCo^*

ic.o.^ -

74,2

U-C--

|Fc^*

11 coo-

54,6

6S.4

CItjCOO-

40.9

iUF'

69.7

CHjCH.COO-

.35.8

NMc*'

44.9

iFctCN)?-

IFc(CN)/

100,9

NEi^ ^

32,7

1 10.5

XV, 2. I’llE MEASUREMENT OF CONDUCriVlT Y. To measure the
conductivity of a solution it is placed in a cell of wliich the cell constant has been
determined by calibration with a solution ofaecuraicly known conductivity, e.g.,
a standard potassium chloride solution. The cel! is placed in one ann of a

616

CONDUCTOMETRIC TITRATIONS XV, 3

Wheatstone bridge circuit as in Fig. XV, 1, and the resistance measured. For
details of the procedure a textbook of practical physical chemistry should be

The passage of a current through a solution of
an electrolyte may produce changes in the
composition of the solution in the vicinity of
the electrodes; potentials may thus arise at the
electrodes, with the consequent introduction of
serious errors in the conductivity measurements,
unless such polarisation elfects can be reduced to
negligible proportions. These difficulties are
generally overcome by the use of alternating
currents for the measurements so that the extent
of electrolysis and the polarisation effects are
greatly reduced. The source of alternating current (V) may be either the electric
mains with a frequency of 50-60 hertz, or a mains-operated oscillator giving
current with a frequency of up to 3000 Hz. Since alternating current is being used,
the cell will have a capacitance which will not be counter-balanced in the
standard resistance box i?, and it is therefore necessary to include a variable
condenser in parallel with the resistance box so that the capacitance in a-c can be
matched in a-b.

If the frequency of the current is greatly increased to 10^-10® Hz,, then the
capacitance and inductive effects become highly important, and the apparatus
must be modified to take account of these effects. It is therefore necessary to
consider separately (a) conductometric titrations carried out with current of low
frequency (up to 3000 Hz), and (b) titrations carried out using current at high
frequencies : in these cases we measure changes in capacitance or in inductance
rather than in conductance, and such titrations are therefore usually referred to
as high frequency titrations.

consulted; see Ref. 2.

Conductometric (low frequency) titrations

XV, 3. THE BASIS OF CONDUCTOMETRIC TITRATIONS. The
addition of an electrolyte to a solution of another electrolyte under conditions
producing no appreciable change in volume will affect the conductance of the
solution according to whether or not ionic reactions occur. If no ionic reaction
takes place, such as in the addition of one simple salt to another (e.g., potassium
chloride to sodium nitrate), the conductance will simply rise. If ionic reaction
occurs, the conductance may either increase or decrease ; thus in the addition of a
base to a sUong acid, the conductance decreases owing to the replacement of the
hydrogen ion of high conductivity by another cation of lower conductivity. This
IS the principle underlying conductometric titrations, i.e.', the substitution of ions
ofone conductivity by ions of another conductivity.

Le^ us consider how the conductance of a solution of a strong electrolyte
A B will change upon the addition of a reagent C*D~, assuming that the
cation A'^ (which is the ion to be determined) reacts with the ion Z)” of the
reagent. If the product of the reaction AD is relatively insoluble or only slightly
ionised, the reaction may be written : . , .

A*B--\-C^£)- =AD + C^B~

617

XV, 3 QUANTITATIVE INORGANIC ANALYSIS

Thus in ihc reaction between A * ions and D~ ions, the A ^ ions arc replaced by
C* ions during the titration. As ilic titration proceeds the conductance increases
or decreases, depending upon whether the conductivity of the C*’ ions is greater
or less than that of the A ^ ion.

During the progress of neutraiisation.s. precipitations, etc., changes in
conductivity may. in general, be e.\pcctcd, aiid these may therefore be employed
in determining the end points a.s well as the progress of the reactions. The
conductivity is measured after each addition of a small volume of the reagent,
and the points thus obtained arc plotted to give a graph which consists of two
straight lines intersecting at the equivalence point. The accuracy of the method is
greater the more acute the angle of intersection and the more nearly the points of
the graph lie on a straight line. The volume of the solution should not change
appreciably ; this may be achieved by employing a titrating reagent whicii is 20 to
100 times more concentrated than the solution being titrated, and the latter
should be a.s dilute as practicable, Thus if the conductivity cell contains about
lOOcm^ of the solution at the beginning of the titration, the reagent (say, of fifty
times the concentration of the solution being analysed) may be placed in a S-cm’’
mieroburette, graduated in 0.0 i or 0.02 cm^. .A correction for the dilution effect
may. however, be made by multiplying the values of the conductivity by the
factor ( F-h \')/V, in which V is the original volume of the solution and v is the
volume of reagent added.

In contrast to potentiometric titration methods (see Chapter XIV), but similar
to amperometric titration methods (sec Chapter XVM), measuremcm.s near the
equivalence point have no special significance. Indeed, owing to hydrolysis,
dissociation, or solubility of the reaction product, the values of the conductivity
measured in the vicinity of the equivalence point are usually worthless in the
construction of the graph, since one or both curves will give a rounded portion at
this point. Even if the conductivity of the reaction product at the equivalence
point is appreciable, tiic reaction may frcqucnily be employed for
conductometric titration if the conductivity of the reaction product AD is
practically completely suppressed by a reasonable e.xccss of .T* or D~. Thus
conductometric methods may be applied where v isuul or potentiometric methods
fail to give results owing to considerable solubility or hydrolysis at the
equivalence point, for c.MimpIe, in many precipitation reactions producing
moderately soluble substances, in the direct titration of weak acids by weak
bases, and in the displacement titration of salts of moderately weak acids or bases
by strong acids or bases. A further important advantage is that the method is as
accurate in dilute as in more concentrated solutions ; it can also be employed with
coloured solutions.

It may be noted that very weak acids, such as boric acid and phenol, which
cannot be titrated potcmiometrically in aqueous solution, can be titrated
conductometrically with relative ea.se. Mixtures of certain acids can be titrated
more accurately by conductometric than by potentiometric (pH) methods. Thus
mixtures of hydrochloric acid (or any other strong acid) and acetic acid (or any
other weak acid of comparable strength) can be titrated with a weak base (e.g..
aqueous ammonia) or with a strong base (e.g., sodium hydro.xide): reasonably
satislactory end-points are obtained.

Attention is directed to the importance of temperature control in conductance
measurements. While the use of a thermostat is not essential in conductometric
titrations, constancy of temperature is required but it is usually only necessary to

618

CONDUCTOMETRIC TITRATIONS XV, 4

place the conductivity cell in a large vessel of water at the laboratory
temperature.

The relative change of conductivity of the solution during the reaction and
upon the addition of an excess of reagent largely determines the accuracy of the
titration ; under optimum conditions this is about 0.5 per cent. Large amounts of
foreign electrolytes, which do not take part in the reaction, must be absent, since
these have a considerable effect upon the accuracy. In consequence, the
conductometric method has much more limited application than visual,
potentiometric, or amperometric procedures.

XV, 4. APPARATUS AND MEASUREMENTS. A conductivity cell for
conductometric titrations may be of any kind that lends itself to thorough
stirring of the contents (preferably by mechanical means), and permits the
periodical addition of reagents. As explained above, it may be necessary to place
the cell in a large vessel of water in order to maintain constancy of temperature,
but in most circumstances the cell may be used at the ambient temperature of the
laboratory. The cell should be constructed of Pyrex or other resistance glass and
fitted with platinised platinum electrodes; the platinising helps to minimise
polarisation effects. The size and separation of the electrodes will be governed by
the change of conductance during the titration : for low-conductance solutions
(e.g., when extremely dilute), the electrodes should be large and close together;
for precipitation reactions the electrodes must be vertical.

The following procedure may be used for platinising the electrodes. The
conductivity vessel and electrodes are thoroughly cleaned by immersion in a
warm solution of potassium dichromate in concentrated sulphuric acid. After
washing with distilled water until free from acid, the electrodes are plated from a
solution containing 3 g chloroplatinic acid and 0.025 g lead acetate per lOOcm^.
The current may be obtained from two accumulators (4 volts), the poles of which
are connected to the ends of a suitable sliding resistance. The current is adjusted
so as to produce a moderate evolution of hydrogen. Each electrode should be
used alternately as anode and cathode (i.e., the current should be reversed every
half minute) and electrolysis should be continued until both electrodes are
covered with a jet-black deposit. The time may vary from about two to about five
minutes. After platinising, the electrodes must be freed from traces of chlorine;

dilute sulphuric acid is electrolysed during
15 minutes using the two platinised elec-
trodes (connected together) as cathode and
another platinum electrode as anode. The
electrodes are then washed with distilled
water and afterwards kept immersed in
distilled water until required for use.

A cell suitable for conductometric ti-
tration is depicted in Fig. XV, 2(a); the
electrodes are firmly fixed in the perspex lid
which is provided with openings for the
stirrer and the jet of the burette. The stirrer
shown may be replaced by a magnetic
stirrer. Alternatively, a three- or four-
necked flat bottom flask may be used, with
the tubes carrying the electrodes fitting into

Fig. XV, 2

s_a

V y

(b)

619

XV, 4 QUANTITATIVE INORGANIC ANALYSIS

ground glass joints and thus being accurately located: with the ihrce-neckcd
flask a magnetic stirrer must be used.

For most purposes a special cell is not required and good results are obtained
by clampingacommercially available dip cell (shown diagrainatically in Fig. XV,
2(/))) inside a beaker which is placed on a magnetic stirrer. With this
arrangement, the dipping cell should be lifted clear of the solution after each
addition from the burette to ensure that the liquid between the electrodes
becomes thoroughly nii.xed. Since absolute conductivity values arc not required
it is not necessars’ to know the cell constant.

The conductance nieasurcmentsare made using a Wheatstone bridge circuit as
c,xp!aincd in Section .XV, 2, and the MEL Conductivity Bridge type 7566/3 (Pye
Unicam Ltd) is a very suitable instrument for this purpose- It is basically a mains-
operated Wheatstone bridge working at the mains frequency (50 Hz) for the
measurement of very low conductivities (0.1~10/jScm~M. but from a built-in
oscillator of frequency 2.9 kHz for iiigher conductivities: the bridge incorporates
an electronic ‘magic eye’ b.ilance point indicator. A range switch selects any one
of a .serie.s of standard resistances, and these provide standard conductance
values of 1. 100, and lOOOOpSand of 1 S, whilst the main dial ofthe bridge moves
over a scale which is directly calibrated to give the ratio of the two arms b-dind
d~c (see E'ig. XV, I). Hence, for any setting of the main dial, the observed
conductance is given by (stand.ird conductance :< scale reading).

To check the correct operation of the bridge, the instrument is set at the
‘calibrate’ (CALl position in whicii two equal resistances are connected across
the bridge arms so that tlie main dial balances in the central ‘unity’ position; it
will be found that in the unbalanced condition the whole .irea ofthe magie-cye
indicator will fluoresce brightly, but on rotating the dial a po-siiion will be found
(exactly over ilic centre divi.viun marked ‘1’) where the area of fluorescence
eoniract.s to a minimum.

To measure tiie conductance of a solution, the l;itter is placed in a suitable
conductivity cell, or a dip cell (Fig. XV, 2(M) is supported in tite solution, and
then connected to the TEST terminals ofthe conductivity bridge. The selector
switch is .set to the appropriate conductance range, and the dial is rotated until a
balance is indic.iicd on the magic eye. The conductivity may be calculated by-
multiplying the observed conductance by the cell constant.

It is also possible to use the bridge with an e.xiernal standard resistance box:
this is connected acro.ss the lenninals marked ‘STD’ and the selector switch is set
to an intennediale position, marked by a red spot, in which the internal standards
are disconnected. In this mode of operation, a resistance R selected in the
resistance box corre.spond.s to a sttmdard conductance of 1 / A’, and the unknown
conductance is then given by (1/A reading on main scale).

For conduetomcirie titnitions a eonveniem procedure is lo use an external
resistance box as de.seribcd above, to set the dial of the bridge to the central
position of the scale, and then to adjust the re.sistance box until balance is
attained. This proce.ss is repeated after each addition of litrant and the recorded
resistance values are then plotted against the volume of titrant. This produces a
curve which is a mirror-image inversion ofthe usual conductance v-v. volume of
titrant curve, but is equally satisfactory for determining the end point.

A number of other conductance bridges are commercially available also
operating on the Wheatstone bridge principle, but the Wayne-Kerr bridge
(Wayne-Kerr Laboratories Ltd) i.s a transformer ratio-arm bridge which is less

620

CONDUCTOMETRIC TITRATIONS XV, 5

affected by stray capacitance than is a Wheatstone bridge (Ref. 3).

XV, 5. APPLICATIONS OF CONDUCTOMETRIC TITRATIONS. Some
typical conductometric titration curves are collected in Fig. XV, 3, a-h.

Strong acid with a strong base. The conductance first falls, due to the
replacement of the hydrogen ion (conductivity 350, Table XV, 1) by the added
cation (conductivity 40-80) and then, after the equivalence point has been
reached, rapidly rises with further additions of strong alkali due to the large
conductivity of the hydroxyl ion (198). The two branches of the curve are straight
lines provided the volume of the reagent added is negligible, and their
intersection gives the end-point (curve (a)). This titration is of practical interest
when the solutions are dark or deeply coloured or if they are very dilute
(lO'^-lO""^ At) ; in the latter case carbon dioxide must be excluded.

Strong acid with a weak base. The titration of a strong acid with a
moderately weak base (K ca. 10“^) may be illustrated by the neutralisation of
dilute sulphuric acid by dilute ammonia solution (curve (b)). The first branch of
the graph reflects the disappearance of the hydrogen ions during the
neutralisation, but after the end-point has been reached the graph becomes
almost horizontal, since the excess aqueous ammonia is not appreciably ionised
in the presence of ammonium sulphate.

Weak acid >vith a strong base. In the titration of a weak acid with a strong
base, the shape of the curve will depend upon the concentration and the
dissociation constant K of the acid. Thus in the neutralisation of acetic acid

= 1.8 X 10“^) with sodium hydroxide solution, the salt (sodium acetate)
which is formed during the first part of the titration tends to repress the ionisation
of the acetic acid still present so that its conductance decreases. The rising salt
concentration will, however, tend to produce an increase in conductance. In
consequence of these opposing influences the titration curves may have minima,
the position of which wilt depend upon the concentration and upon the strength
of the weak acid. As the titration proceeds, a somewhat indefinite break will
occur at the end point, and the graph will become linear after all the acid has been
neutralised. Some curves for acetic acid-sodium hydroxide titrations are shown
in diagram (c).

For moderately strong acids (X ca. 10"^) the influence of the rising salt
concentration is less pronounced, but, nevertheless, difficulty is also experienced
m locating the end-point accurately. Thus curve 1 in diagram id) is obtained upon
titrating 0.005M o-nitrobenzoic acid with 0. 1 30M-potassium hydroxide; the
neutralisation line is slightly curved in the neighbourhood of the end point. There
are two procedures for determining the end-point in the titration of weak acids
with bases. The acid is first titrated with aqueous ammonia solution ; if the end-
point cannot be obtained from this curve with the desired accuracy, a second
titration is carried out using potassium hydroxide solution of the same
concentration. The two curves are practically identical up to the neutralisation
point, and beyond this straight lines are obtained in both titrations, the
intersection of which gives the end-point. In diagram (d), curve 2 is obtained with
0.130M-aqueous ammonia solution. If the end-point is required with great
accuracy, a correction should be applied for the fact that the conductance of the
ainmonium salt is approximately 0.6 per cent lower than that of the, potassium
salt, and the point of intersection should therefore be found when the final
section of curve 2 is raised by this amount.

621

XV, 4 QUANTITATIVE INORGANIC ANALYSIS

ground glass joints and thus being accurately located: with the three-nccked
flask a magnetic stirrer must be used.

For most purposes a special cell is not required and good results are obtained
by clamping a commercially uvailabicdip cell (shown diagrumaiically in Fig. XV,
2(b)) inside a beaker which is placed on a magnetic stirrer. With this
arrangement, the dipping cell should be lifted clear of the solution after each
addition from the burette to emsure that the liquid between the electrodes
becomes ihorouglily mixed, Since absolute conductivity values are not required
it is not necessary to know the cell constant.

The conductance measurements arc made using a Wheatstone bridge circuit as
explained in Section XV, 2, and the MI£I. Conductivity Bridge type 7566/3 (Pye
Unicam Ltd) is a very suitable instrument for liiis purpo.ve. It is basically a mains-
operated Wheatstone bridge working at the mains frequency (50Hz) for the
measurement of very- low conductivities (O.^-lO/zScm'*), but from a built-in
oscillator of frequency 2.9 kliz for higher conductivities; the bridge incorporates
an electronic ‘magic eye' balance point indicator. A range .switch .selects any one
of a scries of standard resistances, and these provide standard conductance
values of 1 , 100, and 10000 fiS and of 1 S. wliil.st the main dial of the bridge moves
over a scale which is directly calibrated to give the ratio of the two anns /^i:/and
(J-c (see Fig. XV‘, 1). Hence, for any .setting of tlic main dial, the observed
conductance is given by (standard conductance x scale reading).

To check tlic coriect operation of the bridge, the instrument is set at the
‘calibrate' (CAL) position in wiiich two equal resistances are connaTcd acros-s
the bridge arms so that the main dial balanec.s in the centrai ‘unity’ position; it
will be found tliat in the unbalanced condition the whole area of the magic-eye
indicator will fluoresce brightly, but on rotating tiic dial a position will be found
(exactly over the centre division marked ‘D where the area of fluorescence
contracts to a minimum.

To measure the conductance of a solution, liic latter is placed in a suitable
condueliviiy cell, or a dip ceil (Fig. XV, 2(h)) is supported in the solution, and
then connected to the TEST terminals of the conductivity bridge. Thu .selector
switch is set to the appropriate conductance range, and tlic dial is rotated until a
balance is indicated on tlio magic eye. The condueliviiy may be calculated by
multiplying the observed conductance by liie cell constant.

It is also po.ssib!c to use the bridge with an external standard resistance box:
this is- connected across the terminals mtirked ‘STD' and the .selector switch is set
to tin iniermediaie position, marked by a red spot, in which the internal standards
are disconnected. In this mode of ojveratioiJ. a rcsi.siance R selected in the
resistance box corresponds to .t .standard conductance of 1/A’, and the unknown
coit^diiciance is then given by ( 1/ A x reading on main scale).

For conductometric titrations a cunveiiicni procedure is to use an external
resistance box as de.scribed aimve. to set the dial of the bridge to the central
position o( tlie scale, and then to adjust the resistance box until balance is
attained. TIiLs process is repeated after each addition of liirant and the recorded
resistance values are then plotted against the v olume of tilrant. This produces a
curve which is a mirror-image inversion of the usual conductance vs. volume ot
titrant curve, but is equally satisfactory for determining the end point.

A number of other conductance Ivridgcs are comitiercially available also
operating on the Wheatstone bridge principle, but the Wayne-Kcrr bridge
(Waync-Kerr Laboratories Ltd) is a transformer ratio-arm bridge which is less

020

CONDUCTOMETRIC TITRATIONS XV, 5

affected by stray capacitance than is a Wheatstone bridge (Ref. 3).

XV, 5. APPLICATIONS OF CONDUCTOMETRIC TITRATIONS. Some
typical conductometric titration curves are collected in Fig. XV, 3, a-h.

Strong acid ivith a strong base. The conductance first falls, due to the
replacement of the hydrogen ion (conductivity 350, Table XV, 1) by the added
cation (conductivity 40-80) and then, after the equivalence point has been
reached, rapidly rises with further additions of strong alkali due to the large
conductivity of the hydroxyl ion (198). The two branches of the curve are straight
lines provided the volume of the reagent added is negligible, and their
intersection gives the end-point (curve (a)). This titration is of practical interest
when the solutions are dark or deeply coloured or if they are very dilute
(10"^-10“‘’’iU) ; in the latter case carbon dioxide must be excluded.

Strong acid with a weak base. The titration of a strong acid with a
moderately weak base (K ca. 10“^) may be illustrated by the neutralisation of
dilute sulphuric acid by dilute ammonia solution (curve (^)). The first branch of
the graph reflects the disappearance of the hydrogen ions during the
neutralisation, but after the end-point has been reached the graph becomes
almost horizontal, since the excess aqueous ammonia is not appreciably ionised
in the presence of ammonium sulphate.

Weak acid with a strong base. In the titration of a weak acid with a strong
base, the shape of the curve will depend upon the concentration and the
dissociation constant K of the acid. Thus in the neutralisation of acetic acid
{Ka= 1.8xl0~^) with sodium hydroxide solution, the salt (sodium acetate)
which is formed during the first part of the titration tends to repress the ionisation
of the acetic acid still present so that its conductance decreases. The rising salt
concentration will, however, tend to produce an increase in conductance. In
consequence of these opposing influences the titration curves may have minima,
the position of which will depend upon the concentration and upon the strength
of the weak acid. As the titration proceeds, a somewhat indefinite break will
occur at the end point, and the graph will become linear after all the acid has been
neutralised. Some curves for acetic acid-sodium hydroxide titrations are shown
in diagram (c).

For moderately strong acids {K ca. 10“^) the influence of the rising salt
concentration is less pronounced, but, nevertheless, difficulty is also experienced
m locating the end-point accurately. Thus curve 1 in diagram {d) is obtained upon
titrating 0.005M o-nitrobenzoic acid with 0. 1 30M-potassium hydroxide; the
neutralisation line is slightly curved in the neighbourhood of the end point. There
are two procedures for determining the end-point in the titration of weak acids
With bases. The acid is first titrated with aqueous ammonia solution ; if the end-
point cannot be obtained from this curve with the desired accuracy, a second
titration is carried out using potassium hydroxide solution of the same
concentration. The two curves are practically identical up to the neutralisation
point, and beyond this straight lines are obtained in both titrations, the
intersection of which gives the end-point. In diagram (d), curve 2 is obtained with
0.130Af-aqueous ammonia solution. If the end-point is required with great
accuracy, a correction should be applied for the fact thatthe conductance of the
armonium salt is approximately 0.6 per cent lower than that of the potassium
salt, and the point of intersection should therefore be found when the final
section of curve 2 is raised by this amount. '

621

XV, 5 QUANTITATIVE INORGANIC ANALYSIS

aqueous ■imninni i commenced with a small amount

and is then H (^“^“■''‘^*>1. say, to neutralise one-third of the acit

concentr ition \ n, ^ f hydroxide solution of the san

concentration. A typical curxe lor 0.005A/-mandclic acid is shown in diagram (r

622

CONDUCTOMETRIC TITRATIONS XV, 5

When all the acid (M) has been neutralised the conductance of the mixture falls
owing to the replacement of the ammonium ion by the less coj^ucting sodium
ion {(NH 4 '^+M“)+(Na+ +OH"),= NH3+.H20 + Na+ +M~}; when the
displacement of the ammonia is complete, the conductance rises abruptly. At
this end-point (S), the, total amount of sodium hydroxide solution added (QS)
is equivalent to the acid originally present (FQ represents the first stage of the
titration performed with aqueous ammonia solution). Alternatively, the acid
present is also measured by the total alkali added to the point (i?) at which the
conductivity falls, i.e., by (FF). A double check is thus obtained in the titration.
The method may be employed to improve the end-point of any titration if the
acid is sufiiciently strong to form an ammonium salt.

Very weak acid rvith a strong base. The initial conductance is very small,
but increases as the neutralisation proceeds owing to the salt formed. The
conductance values near the equivalence point are high because of hydrolysis ;
beyond the equivalence point the hydrolysis is considerably reduced by the excess
alkali. To determine the end-point, values of the conductance considerably
removed from the equivalence point must therefore be used for extrapolation.
Some titration curves for boric acid and sodium hydroxide solution are given in
diagram (f).

Weak acids with weak bases. The titration of a weak acid and a weak base

can be readily carried out, and frequently it is preferable to employ this
procedure rather than use a strong base. Curve (g) is the titration curve of
0.003M-acetic acid with 0.0973M-aqueous ammonia solution. The neutrali-
sation curve up to the equivalence point is similar to that obtained with sodium
hydroxide solution, since both sodium and ammonium acetates are strong
electrolytes; after the equivalence point an excess of aqueous ammonia solution
has little effect upon the conductivity, as its dissociation is depressed by the

ammonium salt present in the solution. The advantages over the use of strong
alkali are that the end-point is more easy to detect, and in dilute solution the
influence of carbon dioxide may be neglected.

Mixture of a strong acid and a weak acid with a strong base. Upon adding a
strong base to a mixture of a strong acid and a weak acid (e.g., hydrochloric and
acetic acids), the conductance falls until the strong acid is neutralised, then rises
as the weak acid is converted into its salt, and finally rises more steeply as excess
of alkali is introduced. Such a titration curve is shown as S in diagram (h). The
three branches of the curve will be straight lines except in so far as ; (a) increasing
dissociation of the weak acid results in a rounding off at the first end point, and
(h) hydrolysis of the salt of the weak acid causes a rounding off at the second end
point. Usually, extrapolation of the straight portions of the three branches leads
to definite location of the end-points. Here also titration with a weak base, such
as aqueous ammonia solution, is frequently preferable to strong alkali for
reasons already mentioned in discussing weak acids : curve W in diagram (/i) is
0 amed by substituting aqueous ammonia solution for the strong alkali. The
procedure may be applied to the determination of mineral acid in vinegar or

other weak organic acids 10 ~^).

tit (or replacement) titrations. When a salt of a weak acid is

ra e with a strong acid, the anion of the weak acid is replaced by that of the
u itself is liberated in the undissociated form,

the addition of a strong base to the salt of a weak base, the cation of

ea base is replaced by that of the stronger one and the weak base itself is

623

XV, 5 QUANTITATIVE INORGANIC ANALYSIS

liberated in the undissociated Ibrni. If, for example, ^/-hydrochloric acid is
added to a 0.1 Af-solution of sodium acetate, the curve shown in Fig, XV, 4 is
obtained; the acetate ion is replaced by the chloride ion. The initial increase in
conductivity is due to the fact that the conductivity of the chloride ion is .slightly

greater than that of the acetate ion. Until
the replacement is nearly complele, the
solution contains suflicient sodium acet-
ate to suppress the ionisation of the
liberated acetic acid and thereby render
negligible it,s contribution to the con-
ductivity of the solution. Near the
equivalence point the acetic acid is
sulltciently ionised to alfcci the con-
ductivity. thus leading to higher values of
the conductivity and the rounded por-
tion of the curve. Beyond the equivalence
point when excess of hydrochloric add is
present, the ionisation of the acetic acid
cm' til tit 1 1 ( s.iOit is again suppressed and the conductivity
rises rapidly. It can ca.sily be calculated
’ that to titrate a 0. l.V/-.salt .solution the

dissociation constant must not be greater
than 5 x for a 0.01.l/-salt solution A’ i- 5 x 10' and for a O.OOl.U-salt
solution K ■{- 5 x lO"". i.e., the ionisation constant of the displaced acid or base
divided by tlie original concentration of the salt must not exceed about 5x10"'’.
Fig. XV, 4 also includes the titration curve of 0.01 ,\/-ammoniuin chloride
solution with O.l.U-sudium hydroxide solution. The dccre.tse in conductivity
during tiic di,splacemcni is caused by the subsiituiiutt ol the ammonium ion by
the sodium km isec Table XV, 1 in Section XV, 11.

Frccipitatiun and complex formation reactions. .A reaction may Ik' made
the basis ofa eonductometric mnnion if the reaction product is sparingly soluble
or is a stable complex, 'fhe following factors must be considered in connection
with the Usefulness and accuracy of the titration;

1. In order to reduce the mtlucncc of errors in the conductometric titration to
a minimum the angle between tlie two branches of tlie curve should be as small as
possible. If the angle is very obtuse, a small error in the conductance data can
cause a large deviation. The following appro.vimate rules will be found useful ;
(a) The smaller the conductivity of the ion which replaces the reacting ion, the

more accurate will be the result. (Thus it is preferable to titrate a silver salt
with lithium chloride rather than with liydrochloric acid; cations may be
titrated with lithium salt.s. and anion.s with acetates.)

(/)) The larger the conductivity of the anion of the reagent which reacts with the
cation to be determined, or vice versa, the more ;icutc is the angle.

(f) The titration of :i slightly ionised sail does not give good results, since the
conductivity increases continuously from the commencement. Ucncc the salt
present in the cell should be virtually completely di.ssociated: for a similar
reason, the added reagent should also be a strong electrolyte.

2. The solubility of the precipitate (or the dissociation of thecomple.x) should
be less iltun 5 per cent. The addition of ethanol is sometimes rcconunendcd to
reduce the solubility, but its inlluencc on the factors detailed in 1 must be borne in

624

CONDUCTOMETRIC TITRATIONS XV, 6

Fig. XV, 5

mind. An experimental curve is given in Fig. XV, 5 (ammonium sulphate in
aqueous-ethanol solution with barium acetate). If the solubility of the precipitate
were negligibly small, the conductance at the equivalence point would be given

by AB and not the observed AC.
The addition of excess of reagent
depresses the solubility of the pre-
cipitate and, if the solubility is not
too large, the position of the point B
can -determined by continuing the
straight portions of the two arms of
the curve until they intersect.

3. A slow rate of precipitation,
particularly with micro-crystalline
precipitates, prolongs the time of
titration. Seeding or the addition of
ethanol (concentration up to 30-40
per cent) may have a favourable

cm^ of reagenl effect,

4. If the precipitate has pro-
nounced adsorptive properties, the

composition of the precipitate will not be constant, and appreciable errors may
result. Occlusion may take place with micro-crystalline precipitates.

In spite of the obvious limitations of the method, quite a large number of
precipitation titrations have been carried out; thus silver nitrate, lead nitrate,
barium acetate or barium chloride, uranyl acetate, lithium sulphate, and lithium
oxalate have been employed in precipitation reactions (for details, see Selected
bibliography— Section XV, 12).

Oridation-reduction (redox) titrations. The conductometric method is not
well suited to the study of oxidation-reduction titrations. Almost all such
reactions must be carried out in the presence of a large excess of acid or base,
which more or less completely masks the change in conductance due to the redox
^ example is the titration of iron(II) with permanganate in

which, say, a 0.01.4/ solution of iron(II) in 0.54/-sulphuric acid is titrated with
• 3/-potassium permanganate. Although the reaction

5Fe^+ + MnO^- -f 8H + = 5Fe^ + -f Mn^ + 4- 4 H 2 O

does consume hydrogen ions, thus decreasing the conductance of the solution up
s° if point, the fraction of hydrogen ion thus removed is relatively

roa . he entire change in conductance is not great and cannot be detected with
curacy by the usual equipment employed for conductometric titrations.

• fOME EXPERIMENTAL DETAILS FOR CONDUCTOMETRIC
• ^UNS. One of the cells described in Section XV, 4 is used in
dip celWF^ ^ Wheatstone bridge operating with alternating current. The
stirrin ' clamped inside a beaker which is set up for magnetic

made ^ If/'^^^^^^cnded, and the conductance measurements are conveniently
conveif^ ” MEL Conductivity bridge. As explained in Section XV, 4, it is
to adiu* bridge in conjunction with an external resistance box, and

point ^ bridge set to unity, to find the balance

625

XV, 7 QUANTITATIVE INORGANIC ANALYSIS

After measuring the solution to be titrated (up to 25 cm^) into the beaker, it is
diluted with distilled water to at least 100 cm^. and the stirrer set in motion. The
titrating agent (concentration at least 10 times that of the solution being titrated)
should be placed in a 5 or 10-cm-' micro-burette; the reagent is added in small
portions, and the solution is stirred or shaken after each addition. The
conductivity is measured after the wetl-mi.scd solution has been allowed to stand
for a minute or two. The addition of the titrating reagent is continued until at
least five readings beyond the equivalence point have been made. It is often
advisable to carry- out a preliminary titration: this will provide information as to
the increments of the reagent best suited for the particular titration, e.g., in
increments of 0.5 cur*, etc. The conductance (or resistance) is plotted as ordinates
against the volume of the titrating reagent as abscissiic; the two straight portions
of the curve are extrapolated until they intersect and the point of intersection is
taken as the equivalence point of the reaction.

In order to obtain satisfactory results by extrapolation procedures from the
titration curves, the following points must be borne in mind. Firstly,
extrapolation can only be performed satisfactorily with lines. Curvature

in the immediate neighbourhocKl of the end point as show n in curved .VF. 3, (e),
(J) and (/i) may be ignored, presided there arc adequate numbers of readings on
both sides of the end-point to permit the drawing of uneciuivocal straight lines:
this requires that sullicient readings be taken after the equivalence point has been
reached to give at least five points' lying on a straight line.

Secondly, it is most important that thedilutioncorrcction factor (F+ v)/F(sce
Section XV, 3) be applied to the readings before plotting the curves.

High frequency litrations

XV, 7. GENERAL CONSIUERATIONS. In the high-frequency method of
titration a suitable cell containing the chemical system is made part of, or is
coupled to, an oscillator circuit resonating at a frequency of several megahertz.
As the composition of the chemical system changes the resistance (impedance)
and/or capacitance of the circuit are altered, and changes arc produced in
oscillator ciiaracterislics, such as frequency, grid current and voltage, and plate
current and voltage. Any of these quantities may be measured and taken as an
indication of the change in composition of thcchcmical .system, e.g., as a solution
is titrated with an appropriate reagent : curves may be generally obtained which
show' inllexion or breaks at the equivalence point. The fundamental properties of
the chemical system which alfect the oscillator characteristics are the dielectric
constant and the conductivity. .An important advantage of the high frequency
method is that the electrodes may be placed on the outside of the cell, and are
therelore not in direct contact with the lest solution. Measurements can
accordingly be made without danger of electrolysis or electrode polarisation;
I urthermore, errors resulting from the coating of the electrodes by a precipitate
and other surface phenomena are eliminated. One disadvantage is that the
response of a high-frequency titrimeter is non-specific, being dependent only on
the conductivity and dielectric constant of the system and is independent of the
chemical identity of the components of the system.

It is interesting to consider what happens to the individual ions of an
electrolyte and to polar molecules when expo,sed to a rapidly alternating field.

626

CONDUCTOMETRIC TITRATIONS XV, 7

Each ion or dipolar molecule tends to move or align itself in the direction of the
electrode of opposite polarity. The electrode polarity changes once every cycle,
and the ion or dipole must reverse its motion or orientation. The conductance of
the solution is the result of the movement of negative and positive ions relative to
their neighbours and the solvent molecules. Each ion tends to move ahead of its
ionic atmosphere, and in consequence an urisymmetrical charge distribution
forms around each central ion, and exerts a retarding force on the ion in a
direction opposite to its motion. At alternating frequencies greater than one
megahertz the central ion changes its direction of motion so rapidly with every
cycle of the applied field that there is little chance for dissymmetry of the ionic
atmosphere to arise, hence the conductance increases. Also at high frequencies
the ions undergo such small oscillations that the oppositely charged ionic
atmosphere exerts a relatively much smaller drag than at low frequencies. Since
the apparent dielectric constant and also the time of formation and decay of the
ionic atmosphere (relaxation time) are both concentration dependent, the
response curves of high-frequency instruments are greatly influenced by ionic
concentration. Indeed, with oscillators which operate at frequencies of the order
of about 2 megahertz, the maximum concentration of electrolyte that can be
employed is of the order of O.OIM The high-frequency technique is most
sensitive in titrations where the total concentration of dissolved ions changes,
e.g., in precipitation and complex formation reactions. It is applicable also to
cases where a fast-moving ion is replaced by a slow-moving ion, as in acid-base
titrations.

In ordinary conductance measurements at frequencies of the order of 1000
hertz the influence of cell capacitance is small, but at megahertz frequencies it
plays an important part. It has been shown that high-frequency titration graphs
can be interpreted on the assumption that the solution in the cell behaves like a
capacitor and resistor in parallel in the oscUlator circuit. A simple cell for high
frequency use may consist of two metallic plates fixed to the walls of a rectangular
glass container (Fig. XV, 6(a)). The cell may be represented by the equivalent

Fig. XV, 6

(b)

Circuit, (b), where is the capacitance through the glass container walls, C, is the
capacitance of the solution (liquid), and R is the resistance of the solution. The
resistance of the glass container walls is assumed to be so high that it may be
neg Circuit (h) may be represented by the simpler equivalent parallel
ircuit, (c), and it can be demonstrated by conventional methods of the theory of
alternating eurrents that :

^ K^+a?(q + c^ (1)

627

XV, 8 QUANTITATIVE INORGANIC ANALYSIS

C>-- + w*(QC,--fC/Q

where w - 2;t/, /being the frequency (hertz)» k is tlic low-frequency conductance
of the solution (— 1/R); and 1/R^ the high-frequency conductance. These expres-
sions give the relationship between the low-frequency conductance, the frequency,
and cither the high-frequency conductance (1/ //,) or the capacitance (Cp).

It can be deduced from equation (1) that for a given electrolyte, the optimum
frequency is related to concentration by the expression

KSxJO^-x-

where D is the dielectric constant and concentration is expressed in terms of the
normal (low frequency) conductivity of the solution (k). For an electrolyte such
as sodium chloride, this equation sliows that the appropriate concentrations for
frequencies of 5, .If) and 100 megahertz are respectively 0.0025 5/, 0.014 .1/, and
0.05 jl/, whilst for an electrolyte of appreciably higher conductivity (i.e., acids
and bases), the appropriate concentrations are signijicanlly reduced: thus for
hydrochloric acid the concentrations corresponding to the .same three fre-
quencies are respectively O.OtKldM. ().()03.\/. and 0.0L\/. Since 30 megahertz
represents a practical upper limit for the construction of apparatus which is
reasonably easy to operate, it follows that there are limitations to the
concentration range over which the method may be used.

In practice, the cell portrayed in Fig. XV, 6(n) will consist either of a glass
ve.sscl witli two conducting metal (aluminium, copper, gold or .silver) bands
encircling it as in Fig. XV, 7 {Section XV, 8), or the glass container will simply fit
snugly inside a coil of wire which covers an appreciable lenglli of the cell. The
first type of cell is termed a eapacitathe cell and equatioas (1) and (2) are
applicable to it; the circuit respond-s tochangc.s in both the conductance and the
dielectric con.siant of the solution in the cell. The second type of cell is termed an
inductive cell and the associated circuit responds only to changes in the
conductance of the cell contents. The capacitative type of cell is usually
preferred for tiirimetry.

CONDUCTOMETRIC TITRATIONS XV, 9/10

Fig. XV, 7

position (about 2.5 cm apart) by a Bakelite strip.
The cell is mounted inside a glass vessel, which
inter alia prevents moisture depositing on the
metal strips, etc.; the glass vessel may be filled, if
desired, with paraffin wax, Connection to the
titrimeter is made with a screened cable. The level
of the liquid in the cell should be about 1cm
above the top of the upper metal strip before
measurements are made; the initial volume is
about 35cm^. The liquid in the cell is stirred
mechanically with an efficient glass stirrer;
alternatively, the bottom of the cell may be
flattened and magnetic stirring employed. For
titrations with somewhat larger volumes of
liquid (>60cm^), a Pyrex cell with dimensions
200mm x 38mm is suitable.

Unless the information is given in the in-

struction manual, it is advisable to ascertain the response of the instrument in

relation to electrolyte concentration: each instrument with a given cell shows
maximum response to changes in the compositon of the test solution over a
somewhat limited range of ionic strength. It is also advantageous if the operating
frequency can be varied, as the shape of the titration curve may well alter with
frequency, and for example, an M-shaped curve which is difficult to interpret, may,
by alteration of frequency be converted to a readily interpreted V-shaped curve.

XV, 9. ADVANTAGES OF THE TECHNIQUE. As indicated in Section
XV, 8, a suitable conductivity cell for high frequency measurements can be easily
constructed from glass tubing by contrast with the special cells with expensive
platinum electrodes which are required for normal conductance measurements.
A special advantage is the fact that the electrodes of the high frequency cell are not
m direct contact with the solution under investigation, and thus no electrolysis
occurs and there can therefore be no polarisation. Furthermore, the electrodes
are protected from contamination, e.g., by precipitated solid during precipitation
titrations, and there is no danger of chemical reactions arising due to catalysis by
the metal surfaces. A further advantage is the applicability of the method to
solutions of low concentration, 10~^-10“'’^M.

Drawbacks to the method are (i) the need for adequate electrical screening, (ii)
the fact that because the electrodes are separated from the solution by the non-
conducting walls of the cell the sensitivity is necessarily less than that associated
with normal conductometric measurements, (iii) the limited concentration range
over which the method can be safely used, and (iv) unless the correct frequency is
chosen, the titration plot may produce curves which are not readily interpreted.

TIONq^' EXAMPLES OF HIGH FREQUENCY TITRA-

... solution to be titrated must, if necessary, be appropriately

auuted, so that when using the cell shown in Fig. XV, 7 (Section XV, 8), a 5 or
cm aliquot of the final test solution, further diluted in the cell to a total

liquid level is just over 1 cm above the upper
one/?' ^ ^ concentration which lies within the optimum

a ing range of the titrimeter to be employed : see Section XV, 8. The stirrer is

629

XV, 8 QUANTITATIVE INORGANIC ANALYSIS

C,- + 4.-^: - W

where to = Inf , /being the frequency (hertz), k h the low-frequency conductance
of tlie solution ( = l/R): and 1/i?;. the Jiigh-frcquency conductance. These e.xpres-
sion.s give the relationship bctv\een the low-frequency conductance, thefrequency,
and either the high-frequency conductance (l/Kp) or the capacitance (Cp).

It can be deduced from equation ( 1 ) that for a given electrolyte, the optimum
frequency is related to concentration by the c.\pressiorj

LSx iO\-/c

where D i.s the dielectric constant and concentration is expres.sed in terms of the
normal (low frequency) conductivity of the .soluliun (v). For an electrolyte such
as sodium chloride, this equation shows that the appropriate concentrations for
frequencies of 5, 30 and 100 megaliert/ are respectively 0.(3025 M, 0.014,1/, and
0.053/, whilst for an electrolyte of appmciably higher conductivity (i.e„ acids
and bases), the appropriate concentrations are significantly reduced: thus for
hydrochloric acid the concentrations corrc-sponding to the same three fre-
quencies are respectively 0.(X>06.V/. 0.(M)3.\/. and 0.01. 1/. Since 30 rnegaherti!
represents a practical up^wr limit for the construction of apptrratus which is
reasonably easy to operate, it follows that there are limitations to the
concentration range over whieli the method may be used.

In practice, the cell portrayed in Fig. XV, (i(«) will consist either of a glass
ves.sel with two conducting metal (aluminium, copper, gold or silver) bands
encircling it as in Fig. XV. 7 (Section ,\V, ,S). or the glass container will simply fit
snugly inside a coil of wire which covers an appreciable length of the cell. The
first type of cell is termed a capacifathc cell and equations (!) and (2) are
applicable to it : the circuit responds to changes in both the conductance and the
dielectric con,stant of the solution in the cell. The second type of cell is termed an
inductive cell and the associated circuit responds only to changes in the
conductance of the cell contents. The capacilative tyi>e of ceil is usmdly
preferred for litrimetry.

XV, 8. APPARATUS. Since it.s introduction as an analytical technique in
1946 (Ref. 4) many instruments have been designed for high frequency
tiirimetry: typical e.xamples will be found in the papers listed in Ref. 5. In
addition, a number of coinmcrcial instruments have also been produced; the
Fisher High Frequency Titrimeier (Fisher Scientific Co), the Sargent
Oscillometer (Sargent and Co), the PCL-Lec Titrimeter (Polymer Consultants
Ltd), in each case, changes in ihe composition of the solution contained in the
titration ceil atfects the high frequency conductance and the capacitance of the
cell with a consequent elTcct upon the circuit of which the cell forms a
component; the various instruments ditlcr in the response (voltage, current,
capacitance or inductance) which is observed, and in the way in which the
response is measured. The instruction manual supplied with the instrument in use
must theretore be consulted for the precise operational details.

A salislactory high frequency conductance celt is shown in Fig. XV, 7. It
consists ot a Pyre.x tube, 150 mm x 25 mm, provided witli two 25-mni wide hands
of aluminium sheet (1.5 mm thick) fitted tightly around the lube and held in

628

CONDUCTOMETRIC TITRATIONS XV, 9/10

position (about 2.5cm apart) by a Bakelite strip.
The cell is mounted inside a glass vessel, which
inter alia prevents moisture depositing on the
metal strips, etc.; the glass vessel may be filled, if
desired, with paraffin wax. Connection to the
titrimeter is made with a screened cable. The level
of the liquid in the cell should be about 1 cm
above the top of the upper metal strip before
measurements are made: the initial volume is
about 35cm^. The liquid in the cell is stirred
mechanically with an efficient glass stirrer;
alternatively, the bottom of the cell may be
flattened and magnetic stirring employed. For
titrations with somewhat larger volumes of
liquid (>60cm^), a Pyrex cell with dimensions
200mm x 38mm is suitable.

Unless the information is given in the in-
struction manual, it is advisable to ascertain the response of the instrument in
relation to electrolyte concentration: each instrument with a given cell shows
maximum response to changes in the compositon of the test solution over a
somewhat limited range of ionic strength. It is also advantageous if the operating
frequency can be varied, as the shape of the titration curve may well alter with
frequency, and for example, an M-shaped curve which is difficult to interpret, may,
by alteration of frequency be converted to a readily interpreted V-shaped curve.

XV, 9. ADVANTAGES OF THE TECHNIQUE. As indicated in Section
XV, 8, a suitable conductivity cell for high frequency measurements can be easily
constructed from glass tubing by contrast with the special cells with expensive
platinum electrodes which are required for normal conductance measurements.
A special advantage is the fact that the electrodes of the high frequency cell are not
in direct contact with the solution under investigation, and thus no electrolysis
occurs and there can therefore be no polarisation. Furthermore, the electrodes
are protected from contamination, e.g., by precipitated solid during precipitation
titrations, and there is no danger of chemical reactions arising due to catalysis by
the metal surfaces. A further advantage is the applicability of the method to
solutions of low concentration,

Drawbacks to the method are (i) the need for adequate electrical screening, (ii)
the fact that because the electrodes are separated from the solution by the non-
conducting walls of the cell the sensitivity is necessarily less than that associated
With normal conductometric measurements, (iii) the limited concentration range
over which the method can be safely used, and (iv) unless the correct frequency is
chosen, the titration plot may produce curves which are not readily interpreted.

TIONS^' examples of high frequency titra-

solution to be titrated must, if necessary, be appropriately
awuted, so that when using the cell shown in Fig. XV, 7 (Section XV, 8), a 5 or
cm aliquot of the final test solution, further diluted in the cell to a total
el° cm^ (so that the liquid level is just over 1 cm^ above the upper

odL"^!' ^ ^ concentration which lies within the optimum

P a mg range of the titrimeter to be employed : see Section XV, 8. The stirrer is

Fig. XV, 7

629

XV, 1 1 QUANTITATIVE INORGANIC ANALYSIS

inserted, and after allowing an adequate warm-up period for the instrument, the
initial reading is made in accordance with the operating instructions. The titrant
should have a concentration Ihe to ten times that of the test solution, and small
portions are added to the cell from a niicroburctle, After each addition ofreagent
the tiirimeter is readjusted; il may be found advisable to stop the stirrer whilst
the adjustments are maiie: finally, the instrument readings are plotted against
the volume of titrant added. Some typical results are shown by the curves in
Fig. XV, 8.

Curves (A) and (B) are typical curves for neutralisation reactions: it is
noteworthy that curve (B) siiows two breaks corresponding to the relictions
CO^^ - +H == HCOj - and HCO^' -t-fr - HiO-f-CO,.

Curve (C) is typical of a comple.xation titration, and curves (Dj, (fi) and (F) are
examples of those obtained during precipitation titrations, f lara and West (Ref.
7) recommend that EDT.A titrations should be performed in buffered solutions,
but the Ni-ED I'A titration curve shown (C). was obtained by working in an
unhulfcred medium. The Tli-o.xalate titration (D) was carried out using test and
littam solutions bolli 0.01 .\/ in nitric acid, whilst the O.Ol .\/ lanthanum acetate
reagent for the lluoride titration (Curve E), coutained a trace of acetic acid
(0.5 cin^ per litre). Thc.se curves were obtained using the apparatus de.scribed in
Ref. 5b.

s to ; -i (, X v) ,s •} io It

Voluiiic af IK'l -.oluiwn Vi.>:uii!c lICl v.'!i;ii<iii VWumcaf tiDFA

(.\) NaOUnithlia (») .NujCO, with HCT (Q NF‘withEDTA

(D) Hi*" wilhNaiC.O, (E) F' wiih La{CiHiO,)j (F) Ik^'wilhNaOH
Fig. XV, 8

XV, 11. References

1. M. L. McGlaslum (1*171), Phyiicit-eheniical Qiiuiiiilics iiml Uiiiis. 2nd edn. London;
Royal Institute ofChemisiry,

2. 0. P. Leviu (1973). Ftndiay's I’ractical Pimiail Chemistry. 9lh edn. Harlow;
Longman.

3. D. K. Browning ( 1 969). PUrirmiieiric Methods. London ; ,\lcGr.iw-Uill.

630

CONDUCTOMETRIC TITRATIONS XV, 12

4. F. W. Jensen and A. L. Parrack (1946). ‘Use of High-frequency Oscillators in
Titrations and Analyses.’ Indl. and Engrg. Client., Analytical Edition, 18, 595.

5. (a) J. L. Hall, J. A. Gibson, H. O. Phillips and F. E. Critchfield (1954). ‘The Use of a •
Radiofrequency Oscillator in Student Analysis.’ J. Chem. Ed., 31, 54.

(b) V. Kyte and A. I. Vogel (1959). ‘An Inexpensive High-frequency Titration
Apparatus for General Laboratory Use.’ Analyst, 84, 1004.

(c) F. Kovac's, O. Klug, M. Gombos and F. Farkas (1971). ‘Apparatus for the
Oscillometric Determination of Concentrations.’ Chem. Anal. (Warsaw), 16, 251.

6. C. L. Wilson and D. W. Wilson (1964). Comprehensive Analytical Chemistry. Vol. Ila.
London; Elsevier.

7. R. Hara and P. W. West (1954). ‘High-frequency Titrations Involving Chelation with
EDTA.’ Anal. Chim. Acta., 11, 264.

XV, 12. Selected bibliography

1. W. G. Berl (1956). Physical Methods in Chemical Analysis. Vols. II/III. New York;
Academic Press.

2. H. T. S. Britton (1934). Conductometric Analysis. London; Chapman and Hall.

3. K. Cruse and R. Huber (1957). Hochfrequenztitration. Weinheim ; Verlag Chemie.

4. P. Delahay (1957). Instrumental Analysis. New York; The Macmillan Co.

5. G. W. Ewing ( 1 975). Instrumental Methods of Chemical Analysis. 4th edn. New Y ork ;
McGraw-Hill Book Co.

6. I. M. Kolthoff and P. J. Elving ( 1 959). Treatise on Analytical Chemistry. Part 1 , Vol. 4.
New York; Wiley (Interscience).

7. I. M. Kolthoff and H. A. 'Lz.ilmen(l9A\). pH and Electro Titrations. New York; John
Wiley.

8. J. J. Lingane (1958). Electroanalytical Chemistry. 2nd edn. New York; Interscience.

9. H. J. S. Sand (1941). Electrochemistry and Electrochemical Analysis. Vol. III.
London; Blackie.

10. P. H. Sherrick, G. A. Dawe, R. Karr, and E. F. Ewen (1954). Manual of Chemical
Oscillometry. Chicago ; E. H. Sargent and Co.

11. C. R. N. Strouts, J. H. Gilfillan and H. N. Wilson (1955). Analytical Chemistry. The
Working Tools. Vol. 11. London; Oxford University Press.

12. H. A. Strobel (1973). Chemical Instrumentation. 2nd edn. Reading, Mass.; Addison-
W esley Publishing Co.

13. A. Weissberger (1960). Physical Methods of Organic Chemistry. Vol. I, Part 4. 3rd
edn. New York ; Interscience.

14. F. J. Welcher (1966). Standard Methods of Chemical Analysis. 6th edn. Vol. 3-A. New
York; Van Nostrand.

15. H. H. Willard, L. L. Merritt, and J. A. Dean (1974). Instrumental Methods of Analysis.
5th edn. New Y ork ; Van Nostrand.

CHAPTER XVI VOLTAIWWIETRY

XVI, 1. INTRODUC'riON. Voitanmiclry ii concerned with the .study of
voltage- current -time rel.itionsiiips during electrolysis carried out in a cell where
one electrode is of relatively large surface area, and the other (the working
electrode) has a very small surface area and is often referred to as a micro-
electrode: the technique commonly involves .viudying the inlUience of voltage
changes on the current flowing in the cell. 'Die micro-electrode is usually
constructed of some inert, conducting inaicrial such as gold, platinum or carbon,
and in some circumstances a dropping mercury electrode (D.M.E) may be used;
for this .vpeeial ease the technique is referred to as polarography.

In view of the relative surface areas of the two electrodes, it follows that at the
large au.viliary or counter electrode the current density will be very small, whilst
at the working electrode it may be high. In consequence, the counter electrode is
not readily polarised, and when small currents flow through the cell, the
concentration of the ions m the electrode layer (i.e,, the layer of solution
immediately adjacent to the electrode) remains virtually equal to the
concentration in the bulk solution, and the potential of the electrode is
maintained at a constant value. By Contrast, at the micro-electrode, ilic electrode
lay cr tends to become depleted of the ions being discharged at the electrode, and if
the solution is not stirred, then tiie diflusiun of ions acro.ss the resultant
concentration gradient becomes an important factor in deteriitliiing the
magnitude of the current flowing.

The total current flowing will m fact be equal to the curreiU carried by the ions
undergoing normal electrolytic migration, plus the current due to the dilTu.sion of
ions

wlicre I is the lota! current, the dilfusion current, and the migration current.
There is, hovvev er, a complicating factor m that, in dilute solution, the depletion of
the electrode layer leads to an inerca.se in the resistance of the solution, and thus
to a change in the Ohm's law potential drop (/ x R) in the cell; consequently the
exact potential operative at the electrode is open to doubt. To overcome this, it is
usual to add an exccs.s of an indilTcrent electrolyte to the system (e.g. 0. 1 Af-KCl),
and under these conditions the solution i.s maintained at a low, constant
resistance, whilst the migration current of the species under- investigation
virtually disappears, i.e., / •= /j.

632

-:fa

VOLTAMMETRY XVI, 2

The rate of diffusion of the ion to the electrode surface is given by Pick’s law as

dc _ Dd^c
dt dx^

where D is the diffusion coefficient, c = concentration, t = time, and x = distance
from the electrode surface (see Ref. 1), and the potential of the electrode is
controlled by the Nernst equation

£ = £9+— In
nF

Qqx

%ed

Techniques which come under the general heading of voltammetry and which
will be treated in this Chapter are:

Polarography (d.c. and a.c.).

Anodic stripping voltammetry.

Chronopotentiometry.

Polarography

XVI, 2. BASIC PRINCIPLES. If a steadily increasing voltage is applied to a
cell incorporating a relatively large quiescent mercury anode and a minute
mercury cathode (composed of a succession of small mercury drops falling
slowly from a fine capillary tube), it is frequently possible to construct a
r eproducible current-voltage curv e. The electrolyte is a dilute solution of the
material under examination (which must be electro-active) in a suitable medium
containing an excess of an indifferent electrolyte (base or ground solution, or
supporting electrolyte) to carry the bulk of the current and raise the conductivity
of the solution, thus ensuring that the material to be determined, if charged, does
not migrate to the dropping mercury cathode. From an examination of the
current-voltage curve, information as to the nature and concentration of the
material m ay be o^ iffed~fHevroYskvr-R€L-2). Heyrovsky and Shikata (Ref. 3)
developeda'n apparatus which increased the applied voltage at a steady rate and
simultaneously recorded photographically the current-voltage curve. Since the
curves obtained with this instrument are a graphical representation of the
polansatioffi ortHedropping^leettjod^the apparatus was called a polarograph,
and the records obtained with it, polarograms; the photographic recorder is now
replaced by a pen recorder, and in some circumstances, by an oscilloscope.

The basi c^appaiatusJ or polarographic analysis is depicted in Fig. XVI, 1. The
dropping mercury electroX is here shown as the cathode (its most common
function); it is sometimes referred to as the working or micro-electrode. The
anode is a pool of mercury, and its area is correspondingly large, so that it may be
regarded as incapable of becoming polarised, i.e., its potential remains almost
constant in a medium containing anions capable of forming insoluble salts with
mercury (Cl , 864 ^ ~, etc.); it acts as a convenient non-standardised reference
e ectrode, the exact potential of which will depend upon the nature and the
concentration of the supporting electrolyte. The polarisation of the cell is
erefore governed by the reactions occurring at the ffiopping mercury cathode,
n et and outlet tubes are provided to the cell for expelling dissolved oxygen from
e solution by the passage of an inert gas (hydrogen or nitrogen) before, but not
unng, an actual measurement — otherwise the polarogram of the dissolved

633

XVI, 2 QUANTITATIVE INORGANIC ANALYSIS

oxygen will appear in the current -voltage curve. P is a potentiometer by which
any e.m.f. up to 3 volts may be gradually applied to the cell. S is a shunt for
adjusting the sensitivity of the galvanoineier G appropriate to the nature and
concentration of the substance being investigated. It may be mentioned that
under tlie.se conditions the current-voltage curve is really a current-cathode
potential curve, but displaced by a constant voltage corresponding to the
potential of the anode. For some purposes it i.s advisable to employ an external
anode of known potential {e.g„ a saturated calomel electrode): an internal
electrode is more convenient for most analytical work, since absolute values of
the cathode potential are not usually required.

The initial potential of the dropping mercury cathode is indeterminate, and
will assume any potential applied to it from an external source; when it acquires a
potential dilTereni from that which it had in the absence of electrical connections,
the w orking electrode is said lo be polurisal.

Let us consider what will occur il'aii external e.m.f. i.s applied to the cell shown
in Fig. XVI. 1, charged witli, say. a dilute, oxygen-free solution of cadmium
chloride. All the positively charged ions present iit the solution will be attracted to
the negative working electrode by; (u) an electrical force, due to the attraction of
oppositely charged bodies to each other, and by (/>) a dilfnsivc forccj arisingfrom
tlie coneentration gradient produced at the el^'cirodirs"u'rfa*ce. The total current
pasJing through the cell can be regarded ;is the suni of these two factors. A typical
simple current-voltage curve i.s shown in Fig. XVI, 2, The working electrode.

being perfectly polarisable. assumes the correspondingly increasing negative
potential applied to it; from A to B practicxiily no current will pass through the
cell. At B, where the potential of the micro-electrode Ls equal to the deposition
potential of the cadmium ions with respect to a mclaliic cadmium electrode, the
current suddenly commences to increase and the working electrode becomes
depolari,sed by the cadmium ion.s, which arc then discharged upon the electrode
surface to form metallic cadmium, consequently a rapid increase in the current
flowing through the cell will be observed. At the point C the current no longer

634

VOLTAMMETRY XVr,2

increases linearly with applied potential but approaches a steady limiting value at
the point D : no increase in current is observed at higher cathode potentials unless
a second compound able to depolarise the working electrode is present in the
solution. At any point on the curve between B and C (usually spoken of as the
polarographic wave) the number of cadmium ions reaching the micro-electrode
surface as a result of migration and diffusion from the main bulk of the solution
always exceeds the number of cadmium ions which react at and are deposited
upon the electrode. At the point C the rate of supply of the cadmium ions from the
main bulk of the solution to the working electrode surface has become equal to
the rate of their deposition. Hence at potentials more negative than point D, the
concentration of undischarged cadmium ions at the micro-electrode surface is
negligibly small relative to the cadmium-ion concentration in the bulk of the
solution; no further increase in current passing through the electrolytic cell can be
expected, since the limiting current is now fixed by the rate at which cadmium
ions can reach the electrode surface.

A number of polarisable micro-electrodes (e.g., a rotating platinum wire, ca. 3
mm long and 0.5 mm diameter, suitably mounted, or stationary noble-metal
electrodes) have been used in determining current-voltage curves, but the most
satisfactory is a slowly growing drop of mercury issuing, under a head of 40-60
cm of mercury, from a resistance-glass capillary (0.05-0.08 mm in diameter and
5-9 cm long) in small, uniform drops. The dropping mercury electrode has the
following advantages:

(a) Its surface is reproducible, smooth, and continuously renewed; this is
conducive to good reproducibility of the current-potential curve and
eliminates passivity or poisoning effects.

{b) Mercury forms amalgams (solid solutions) with many metals.

(c) The diffusion current assumes a steady value immediately after each change
of applied potential, and is reproducible.

(d) The large hydrogen overpotential on mercury renders possible the
deposition of substances difficult to reduce, e.g., the alkali metal ions,
aluminium ion and manganese(II) ion. (The current-potential curves of these
ions are inaccessible with a platinum micro-electrode.)

(e) The surface area can be calculated from the weight of the drops.

The dropping mercury electrode may be applied over the range + 0.4 to about
-2.0 volts with reference to the S.C.E. Above -t- 0.4 volt mercury dissolves and
gives an anodic wave; it begins to oxidise to mercury(I) ion. At potentials more
negative than about —1.8 volts vj. S.C.E., visible hydrogen evolution occurs in
acid solutions and the usual supporting electrolytes commence to discharge. The
range may be extended to about —2.6 volts vs. S.C.E. by using supporting
electrolytes having higher reduction potentials than the alkali metals ; tetra-alkyl
anunonium hydroxides or their salts are satisfactory for this purpose.

Reference has already been made to the convenience for routine analytical
work of using a mercury pool covering the bottom of the electrolysis cell as the
non-polarisable reference electrode; the mercury pool is connected to an external
circuit via a platinum wire sealed through the wall of the cell. If the solution
covering the mercury pool contains chloride ion, the mercury pool acts as a
calomel electrode of the particular chloride-ion concentration. Whilst convenient
deV°^^v*^ polarographic determinations, the mercury pool never possesses a
nite, known potential and the potential does not attain a constant value in the

sence of chloride ions or other depolarising ions: further the internal reference

635

XVI, 2 QUANTITATIVE INORGANIC ANALYSIS

oxygen will appear in the current-voltage curve. P k a potentiometer by which
any e.mi. up to 3 volts may be gradually applied to the cell. S i.s a shunt for
adjusting the sensitivity of the galvanometer G appropriate to the nature and
concentration of the substance being investigated. It may be mentioned that
under these conditions the current-voltage curve is really a current-cathode
potential curve, but di.splaceil by a constant voltage corresponding to the
potential of the anode. For some purposes it is advisable to employ an c.xtcnial
anode of known potential (e.g., a saturated calomel electrode): an internal
electrode is more convenient for most analytical wor.k, sitice absolute values of
the cathode potential are not usually required.

The initial potential of the dropping mcreury cathode is indeterminate, and
will assume any potential applied to it from an external source; when it acquires a
potential dilTcrent from that which it had in the absetice of electrical connections,
the working electrode Ls .said to be potariju-il.

Let us consider what will occur if .'111 external e.m.f. is applied to the cell shown
in Fig. ,\VI, 1, ciiarged with, say, a dilute, oxygen-free .solution of cadmium
chloride. All the po.sitive!y charged ions present in thesoluiion will be attracted to
the negative working electrode by: (is) an electrical force, due to the allraclion of
oppositely eliarged bodies to each other, and by (h) a dilfusive for^^irisingfrora
tlie conceniniiion gradient produced at the e!^;ctfode surface. The total current
pa.ssing through the cell can be regarded .is the suni of these two faetons. A typioil
simple current- voltage curxe is shown in Fig. XVJ, 2. 'J'he working electrode.

being perfectly poiarisable, assumes the correspondingly increasing negative
potential applied to it; from A to B practically no current will pass through the
cell. At B, where the potential of the micro-electrode is equal to the deposition
potential of the cadmium ions with respect to a metallic cadmium electrode, the
current suddenly commence.s to incrca.se and the working electrode becomes
depolarised by the cadmium ions, which are then discharged upon the electrode
surface to form metallic cadmium, consequently a rapid increase in the current
flowing through the cell will be ob.scrvcd. At the point C the current no longer

634

VOLTAMMETRY XVI, 3

An example will make this conception of supporting electrolyte clear. Let us
imagine an electrolytic solution is composed of potassium ions O.IOM and
copper(II) ions 0.005M. If we assume that the molar conductivities of K"*" and
jCu^'*' are approximately equal, then it follows that ca. 90 per cent of the current
will be transported to the cathode by the potassium ions and only 10 per cent by
the copper ions. Both ions will tend to diffuse towards any portion of the solution
where a concentration gradient exists, but the rate of diffusion will be slow; If the
concentration of the potassium ions be increased until it represents 99 per cent of
the total cations present, practically all the current passing through the cell will be
transported by the potassium ions. Under such conditions the electro-active
material can reach the electrode surface only by diffusion. It must be emphasised
that the supporting electrolyte must be composed of ions which are discharged at
higher potentials than, and which will not interfere or react chemically with, the
ions under investigation.

Diffusion current When an excess of supporting electrolyte is present in
the solution the electrical force on the reducible ions is nullified; this is because
the ions of the added salt carry practically all the current and the potential
gradient is compressed or shortened to a region so very close to the electrode
surface that it is no longer operative to attract electro-reducible ions. Under these
conditions the limiting current is almost solely a diffusion current. Ilkovic (Ref. 7)
examined the various factors which govern the diffusion current and deduced the
following equation:

1^ = 601 nD^'^

where = the average diffusion current in microamperes during the life of the
drop;

n = the number of faradays of electricity required per mol of the electrode
reaction (or the number of electrons consumed in the reduction of
one mol of the electro-active species);

the diffusion coefficient of the reducible or oxidisable substance
expressed as cm^ sec“ ^ ;
its concentration in millimoles per dm^ ;

the rate of flow of mercury from the dropping electrode expressed in
mg per second; and
drop time in seconds.

The constant 607 is a combination of natural constants, including the Faraday
constant; it is slightly temperature dependent and the value 607 is for 25 °C. The
Ilkovic equation is important because it accounts quantitatively for the many
factors which influence the diffusion current: in particular, the linear dependence
of the diffusion current upon n and C. Thus, with all the other factors remaining
constant, the diffusion current is directly proportional to the concentration of the
electro-active material this is of great importance in quantitative polarographic
analysis.

D =

C =
m =

t =

The original Ilkovic equation in-
curvature of the mercury surfai./'
right-hand side of the equat'*# ' '
and has a value of 39,
usually has a value bet . ’ ’

very accurate work;

The diffusion ■ <> ■

'{ 1/2

e effect on the diffusion current of the
. ty be : 1 • ’• multiplying the

'*^1;. . V is a constant

. .-f yin parentheses

■ account of in

■ = '’•e, the

637

VOLTAMMETRY XVI, 3

applied voltage curve (polarogram) are
shown in Fig. XVI, 4.

The conventional method of drawing
the current-voltage curves is to plot the
applied e.m.f. as abscissae reading in
increasing negative values on the right:
current is plotted as ordinates, cathodic
currents (resulting from reduction) being
regarded as positive and anodic currents
negative. The height of the curve (wave
height) is the diffusion current, and is a
Applied voltage function of the concentration of the
Fig. XVI, 4 reacting material; the potential cor-

responding to the point of inflection of the
curve [half-wave potential) is characteristic of the nature of the reacting material.
This is the essential basis of quantitative and qualitative polarographic analysis.

The underlying theory may be simplified as follows. Polarography is
concerned with electrode reactions at the indicator or micro-electrode, i.e., with
reactions involving a transfer of electrons between the electrode and the
components of the solution. These components are called oxidants when they can
accept electrons, and reductants when they can lose electrons. The electrode is a
cathode when a reduction can take place at its surface, and an anode when
oxidation occurs at its surface. During the reduction of an oxidant at the cathode,
electrons leave the electrode with the formation of an equivalent amount of the
reductantin solution. Similarly, during the oxidation of a reductant at the anode
electrons pass from the solution to the electrode and form an equivalent amount
of the oxidant. Free electrons cannot exist in solution, consequently any process
of reduction at the cathode is accompanied by a simultaneous oxidation. We may
summarise the above discussion by the equation

Oxidant + n Electrons Reductant

Ox + ne:^Red

( 1 )

The reductant differs from the oxidant merely by n electrons, and together they
form an oxidation-reduction system. We will consider the reversible reduction of
an oxidant to a reductant at a dropping mercury cathode. The electrode potential
is given by:

E®+— In^
nF

where a„^ and a,^^ are the activities of the oxidant and reductant respectively as
they exist at the electrode surface (henceforth called the electrode-solution
interface and denoted by the subscript V), R is the gas constant, T the absolute
temperature, n the number of electrons involved in the reaction, and F the
araday constant; is the electrode potential of the system when the activities
0 the oxidant and the reductant are equal. Polarographic measurements are
se om more accurate than + 1 millivplt, hence substitution of concentrations for
ac ivities will not introduce any appreciable error. Equation (2) may therefore be

£ = £©+ In

[Red],

639

XVI, 3 QUANTITATIVE INORGANIC ANALYSIS

Here is the standard potential of the reaction against tlie reference electrode
used to measure the potential of the dropping electrode, and the potential £ refers
to the average value during the life of a mercury drop. Before the commencement
of the polarographic wave only a small residual current Hows, and the
concentration of any electro-active substance must be the same at the electrode
interface as in the bulk of the solution. As soon as the decomposition potential is
exceeded, some of the reducible substance (oxidant) at the interface is reduced,
;md must be replenished from the body of the solution by means ofdilTusion. The
reduction product (reductant) does not accumulate at the interface, but dilTuses
away from it into the solution or into the electrode material. If the applied
potential is increased to a value at which all the oxidant reaching the interface is
reduced, only the newly farmed reductant will be present; the current then
llowing will Ik* the dilTusion current. The current / at any point on the wave is
determined by the rate of liilTusion of the o.xidani from the bulk of the solution to
the electrode surface under a concentration gradient (Ox) to [OxJ,.

/ - A'l[OxI-[Oxl,)* (g)

When [Ox], is reduced to almost zero, equation (4) may be written:

/ A-(Ox] (5)

where /j is the dilTusion current. From equations (4) and (5). it follows that;

{0.xl.-(/."/J.'A (6)

If the reductant (Red) i vspjublu m water and none w as or i einallv present with the
o.xi dant . it will djllvise jVonyjhc'sm^ electrode to the bulk ofiHer6luti^.

The concentration of[Kedj, at the sin lace at any value of / will be proportional to
the rate of dilTusion of the reductant from the surface of the electrode to the
solution (under a concenlraium gradient (Red),) and hence .ilso the current:

/ •■= k(RedJ,* (7)

If the reductant is insoluble m water but .soluble in the mercury phase (amalgam
formation), equ;ition (7) still iiolds. .Substituting in equation (.1), we have:

RT, K
•--In •
lit k

iiF

, £7-
+ In
nl-

h- I
1

(S)

where £'-"=(/■'-*- A’) and A’ % la .

nl- k

When I is equal to equation (S) reduec.s to:

£, , ^ /•■

»F lj/2

(9)

! he potential at the point on the polarographic wave where the current is equal to
one-half tJie dilTusion current is termed the half-wave potential and is designated

* llieconslaiits /v aniiAiii.i) l\:cv.ilii.ilcd frnm the Ilkaviecn'iaUon.

640

VOLTAMMETRY XVI, 3

by £i/ 2 - equation (9) that £^2 is a characteristic constant for

a reversible oxidation-reduction system and that its value is independent of the
concentration of the oxidant [Ox] in the bulk of the solution. It follows from
equations (8) and (9) that at 25° :

0.0591, h-I
£ = £i/ 2 +— — log— j—

( 10 )

This equation represents the potential as a function of the current at any point on
the polarographic wave; it is sometimes termed the equation of the polarographic
wave.

The half-wave potential is also independent of the electrode characteristics,
and can therefore serve for the qualitative identification of an unknown
substance. Owing to the proximity of many different half-wave potentials, its use
for qualitative analysis is of limited application unless the number of possibilities
is strictly limited by the nature of the unknown. The theoretical treatment for
anodic waves is similar to the above.

It follows from equation (10) that when log(/j — /)// (where I is the current at
any point on the polarographic wave minus the residual current) is plotted
against the corresponding potential of the microelectrode (ordinates), a straight
line should be obtained with a slope of 0.059 1/n for a reversible reaction; the
intercept of the graph upon the vertical axis..g ives the half-wave potential of t he
system. Hence n, the number of electrons taking part in the reversible reaction,
may be determined. In applying equation (10), it is necessary to correct both I and
Ij for the residual currents at the corresponding values of the applied potential
and to correct the applied potential itself for any IR drop in the cell circuit. After
these corrections have been made, both £ 1/2 and the slope of the log plot are
found to be independent of the concentration of the electro-active ion. Because it
is concentration independent, the half-wave potential is generally preferred to the
somewhat vague ‘decomposition potential’. It also follows from equation (10)
that the range of potentials over which the polarographic wave extends decreases
with increasing values of n; thus the wave is steeper in the reduction of the
trivalent aluminium or lanthanum ion than forthe lead or cadmium ion, which in
turn is steeper than that of an alkali metal ion or thallium(I) ion.

If the reaction at the indicator electrode involves complex ions, satisfactory
polarograms can be obtained only if the dissociation of the complex ion is very
rapid as compared with the diffusion rate, so that the concentration of the simple
ion is maintained constant at the electrode interface. Let us consider the general
case of the dissociation of a complex ion:

The instability constant may be written:

( 11 )

(st^tly, activities should replkce concentrations).

c may imagine the electrode reaction to be (assuming amalgam formation):

AI"^+ne-^Hg^M(Hg) (13)

Combining ( 11 ) and (13), we have:

+ ne + Hg ^M(Hg) +pX'’- (14)

641

XVI, 4 QUANTITATIVE INORGANIC ANALYSIS

It can fac shown* that the expression for the electrode potential can be written:

. 0.0591 ,

0.0591,

log (A'"

(15)

Here p is the coordination number of the complex ion formed, A'*’' is the ligand
and ;i is the number of electrons involved in the electrode reaction. The
concentration of the complex ion does not enter into equation (15), so that
the ob.served half-wave potential will be constant and independent of the
concentration of tiie complex metal ion. Furthermore, ilic half-wave potential is
more negative the smaller the value of i.e.. the niorestable thecompiex ion.
The halfwave potential will also .shift with a change in the concentration of the
ligand, and if the former is determined at two dilTereni concentrations of the
complex formittg agent, we have;

(16)

This relationship enables one to determine the coordination number p of the
complex ion and thus its formula.

It can also be shown that:

0.0591, „ 0.0591,

(f, 2 !, 0, --- ""■-logA,::,:,s "P-— --- !og(A'’ ]

(17)

where (K, -), and (Fi,.;), are the half-wave potentials of the complex and simple
ions respectively at 25 C, A",.,,,,,, is the instability (or dissociation) constant and
[A''*'] is the concentration of the complexing agent A’*'' in the body of the
solution. It is assumed that ilie ligand is present in sullicicntly large amount so
that its conceuUation is practically the same at the surface of the dropping
electrode a.s in the bulk ofthe solution. This formula may bccmploycd to evaluate
the instability constant of the complex ion: it involves merely the comparison of
the half-wave potential at a given concenUation of the complexing agent with
that of the simple metal ion.

The shift of the half-wave potcnti.ds of metal ions by complexation is of value in
polarographic analysis to eliminate the interfering clfcci of one metal upon
another, and to promote sullicicm separation of the waves of metals in mixtures
to make possible their simultaneous determination. Thus in the analysis of
copper-ba.se alloys for nickel, lead, etc., the reduction wave of coppet(i!) ions in
most supporting electrolytes precedes tliat of tlie other metals and swamps those
of the other metals present; by using a cyanide supporting electrolyte, the copper
is converted into the diilicultly reducible cyanocupratc(l) ion and. in such a
medium, nickel, lead, etc., cun be determined.

XVI, 4. QUANTITAITVE TECll.NIQUE. General coosideratioiis. Polaro-
graphic analysis is mo.sl conveniently carried out if the concentration ot
the electro-active substance is 10' ’‘-lO ■' molar and the volume ofthe solution is
between 2 and 25 cm’. It is, however, possible to deal with concentrations as high
as 10 * molar or as low as 10~’ molar and to employ volumes appreciably less

• Sec jny text-book on Pokirography. c.g. Kef. 4.

642

VOLTAMMETRY XVI, 4

than 1 cm^. Under normal conditions (in particular, concentrations of 0.0001-
0.00 IM) and with strict adherence to.established technique, the reproducibility of
duplicate analyses may be as good as + 2 per cent.

polarographic behaviour is of considerable practical importance. A typical
polarogram for air-saturated M-potassium chloride solution (in. the presence of
0.01 per cent methyl red) is given in Fig. XVI, 5 (curve A). It has been stated that
the first wave (starting at about —0.1 volt relative to S.C.E.) is due to the reduction
of oxygen to hydrogen peroxide; ■ ’

02 + 2H20 + 2e= H202 + 20H~ (neutral or alkaline solution)

O2 + 2H + 2e == Hj O2 (acid solution)

The second wave is ascribed to the reduction of the hydrogen peroxide either to
hydroxyl ions or to water:

H202+2e= 20H" (alkaline solution)

H2O2 + 2H'" -h2e= 2H2O (acid solution)

It is therefore necessary to remove any dissolved oxygen from the electrolytic
solution whenever cathodic regions are being investigated in which oxygen
interferes. This is easily accomplished by bubbling an inert gas (nitrogen or
hydrogen) through the solution for about 10-15 minutes before determining the
current-voltage curve. Curve B in Fig. XVI, 5 was obtained after the removal of
the oxygen by unpurified nitrogen from a cylinder of the compressed gas. The gas

stream must be discontinued dur-
ing the actual measurements to
prevent its stirring effect inter-
fering with the normal, formation
of drops of mercury or with the
diffusion process near the micro-
electrode. Commercial nitrogen or
hydrogen derived from a cylinder
of compressed gas (usually con-
taining less than 0.05 per cent of
oxygen) may be purified by pass-
ing through a Pyrex tube charged
with copper gauze heated to about
450 °C and then through a wash
bottle to saturate it with , water
vapour before passing through the

^ j - • • in the polarographic

cell, the latter procedure minimises any change in volume of the test solution due
to evaporation. It is more convenient, however, to use pure nitrogen (oxygen free!
which can be purchased in cylinders. jb h

The influence of temperature has already been discussed. The electrolytic cell

Edme''®- Mercury pool anode, volts

Fig. XVI, 5

.643

XVI, 4 QUANTITATIVE INORGANIC ANALYSIS

It can be shown* tliai tiie expression for the electrode potential can be written:

_ aostii , ,,

=--£'^ + log K., ah.-

If- II

0.0591

Iog{A-‘’-F

(15)

Here p is the coordination number of the complex ion formed, X‘’~ is the ligand
and n is the number of c!cctron.s involved in the electrode reaction. Tlie
concentration of the complex ion docs not enter into equation (15), so that
the observed half-wave potential will be constant and indei>endent of the
concentration of the complex metal ion. Furthermore, the lialf-vvavc potential is
more negative the smaller the value of , i.e., the more stable thccomplex ion.
The half-wave potential will also shift with a change in the concentration of the
ligand, and if the former is determined at two dilTercnt concentrations of the
complex forming agent, we have:

f.--- -p — — -xAlog[,Y'-)

(16)

This relationship enables one to determine the coordination number p of the
complex ion and thus its fornnila.

It can also be shown that:

^ „ 0.0591, „ 0.0591, ,

(17)

where (Ky,,^ and arc the half-wave potentials of the complex and simple
ions respectively ;ii 25 C, is the instability (or dissociation) constant and
(A''’"] is the concentration of the complexing ;igent A'*" in the body of the
solution. It is a.Ssumcd that the ligand Is present in sutliciently large amount so
that its concentration is practically the same ;it the surface of the dropping
electrode as in the bulk of the solution. This formula may l>e employed to evaluate
the iustahility constant of the complex ion: it involves merely the comparison of
the half-wave potential at a given concentration of the complexing agent with
that of the simple incial ion.

The shift of the half-wave potentials of metal ions by complcxalion is of value in
polarographic analysis to eliminate the interfering clTect of one metal upon
another, and to promote sutlieienl separation of the waves of metals in mixtures
to make possible their sinniltuneous determination. Thus in the analysis of
copper-base alloys for nickel, lead, etc., the reduction wave of coppcr(Il) ions in
most supporting electrolytes precedes iliat of the other metals and swamps those
of the other metals present; by usinga cyanide supporting electrolyte, the copper
i.s converted into the ditficultly reducible cyanocupratell) ion and. in such a
medium, nickel, lead. etc., can be dcicrmincd.

XVI, 4. QUANTI T.VnVE lECUNIQUE. General coiLsiderations. Polaro-
graphic analysis is most conveniently carried out if the concentration of
the clectro-activc substance is 10 ■* - 1 0 ' •* molar and the volume of the solution is
between 2 and 25 cnv*. It is. however, possible to deal with concentrations as high
as 10 - molar or as low as 10” * molar and to employ volumes appreciably less

.See any texi-book on Polarogr.iphy. c.g.. Kef. 4.

642

VOLTAMMETRY XVI, 4

than 1 cm^ Under normal conditions (in particular, concentrations of 0.0001-
0.00 IM) and with strict adherence to.estabfished technique; the reproducibility of

duplicate analyses may be as good as ±2 per cent. ■

Oxygen dissolved in electrolytic solutions is easily reduced at the dropping
mercury electrode, and produces a polarogram consisting of two waves of
approximately equal height and extending over a considerable voltage range,
their position depends upon the pH of the solution, being displaced to higher
voltages by alkali. The concentration of oxygen in aqueous solutions that a.re
saturated with air at room temperature is about 2.5 x 10 *M, consequently its
polarographic behaviour is of considerable practicaT importance. A typical
polarogram for air-saturated M-potassium chloride solution (in the presence of
0.01 per cent methyl red) is given in Fig. XVI, 5 (curve A). It has been stated that
the first wave (starting at about — 0.1 volt relative to S.C.E.) is due to the reduction
of oxygen to hydrogen peroxide:

02+2H20+2e= H202 + 20H~ (neutral or alkaline solution)'

02+2H^ +2e = H202 (acid solution) , •

The second wave is ascribed to the reduction of the hydrogen peroxide either to
hydroxyl ions or to water:

H2O2 + 2e= 20H'" (alkaline solution)

H2O2 +2H'*^ +2e= 2H2O (acid solution)

stream must be discontinued dur-
ing the actual measurements to
prevent its stirring effect' inter-
fering with the normal formation
of drops of mercury or with the
diffusion process near the micro-
electrode. Commercial nitrogen or
hydrogen derived from a cylinder
of compressed gas (usually con-
taining less than 0.05 per cent of
oxygen) may be purified by pass-
ing through a Pyrex tube charged
with copper gauze heated to about
450 °C and then through a wash
bottle to saturate it with water
vapour before passing through the
,, test solution in the polarographic

cell; the latter procedure minimises any change in volume of the test solution due
to evaporation. It is more convenient, however, to use pure nitrogen (oxygen free)
which can be purchased in cylinders. ■ ’

The influence of temperature has already been discussed. The electrolytic cell

Edm£''S- Mercury pool anode, volts

Fig. XVI, 5

XVI, 4 QUANTITATIVE INORGANIC ANALYSIS

It din be shown* that the expression for the electrode potenib! can be written:

n ~ n

(15)

Here p is the coordination number of the complex ion formed, " is the ligand
and II is the number of electrons involved in the electrode reaction. The
concentration of the contplcx ion does not enter into equation (15). so that
the observed half-waxc ix)teniiai will be comstant and independent of the
concentratton of the complex metal ion. Furthermore, the half-waxe potential is
more negative the smaller the value of i.e., the more stable the co.mple.x ion.

The h.ili^wave potential will also shift with a change in the concentration of the
ligand, and if the former is determined at two dilTereni concentrations of the
complex .forming agent, we have:

.r X. fv--l

(16)

This relationship enables one to determine the coordination number p of the
complex ion and thus its formuhu
It can also be shown that;

(^i rh' ~ (^1 ih

0,0591

be A'.,

-p-

0.0591, ,

log .1'’ ]

(17)

where (h'l ;).. and (£, j>, are the half-wave potentials of the compie.v and simple
ions tcspeclivelv at 25 C, A..,:*., is the instabihtv tor dLvsocialion) constant and
' j is the concenttaiion of the complexing agent A'" in the body of the
.solution, it is assumed that the ligand is present in suOieientiy large amount so
that Its concentration is pr.iciically the same .it the surface of the dropping
electrode as in the bulk ofiiie solution. This formula may be cmploved to evaluate
the iu-stahility constant of the complex ion; n involves merely the comparison of
tile h,i!f-wave poicniiai at a given concentration of the coinplexing agent with
that of the simple metal ion.

The shift of the ha!f-wavepolciui.dsofmeia! ions by complexationisofv-alueia
polarographic analysis to elimin.Ue the interfering clTecl of one metal upon
another, and to promote suincient separation of tlic waves of metals in mixtures
to make possible their Minulianeous determination. Thus in the analvsis of
copper-base alloys for nickel, lead, etc., the reduction wave of copper! H) ions in
most supporting eleciroly les precedes that of the other metals and swamps those
of the other metals present; by using a cyanide supporting electrolyte, the copper
is converted into the diliiculily reducible cyaiiocupraieH) ion and, in such a
medium, nickel, lead, etc., can be determined.

XVI, 4. QUA.VriT.ATlVE rECHMQUE. General coasideratioas. Polaro-
graphic analysis is most conveniently carried out if the concentration of
the eleciro-aclivc substance is 10' 10 '■* molar and the volume of the solution is

between 2 and 25 ctnT It is, however, possible to deal with concentrations a.s high
as 10 ■ - molar or as low as 10' * molar and to employ volumes appreciably less

* Sec any tcu-bcoV; on Polaro^raph), c g,. Ref. 4.

642

VOLTAMMETRY XVI, 4

Oxygen dissolved in electrolytic solutions is easily reduced at the dropping
mercury electrode, and produces a polarogram consisting of two waves of
approximately equal height and extending over a considerable voltage range,
their position depends upon the pH of the solution, being displaced to higher
voltages by alkali. The concentration of oxygen in aqueous solutions that are
saturated with air at room temperature is about 2.5 x 10 consequently, its
polarographic behaviour is of considerable practical importance. A typical
polarogram for air-saturated M-potassium chloride solution (in the presence of
0.01 per cent methyl red) is given in Fig. XVI, 5 (curve A). It has been stated that
the first wave (starting at about —0.1 volt relative to S.C.E.) is due to the reduction
of oxygen to hydrogen peroxide:

Oa + 2 H 2 O + 2e = H 2 O 2 + 20H ~ (neutral or alkaline solution)

O 2 + 2H ^ + 2e = H 2 O 2 (acid solution)

The second wave is ascribed to the reduction of the hydrogen peroxide either to
hydroxyl ions or to water:

H202-f2e= 20H” (alkaline solution)

H 2 O 2 -I- 2H ^ + 2e = 2 H 2 O (acid solution)

stream must be discontinued dur-
ing the actual measurements to
prevent its stirring effect inter-
fering with the normal formation
of drops of mercury or with the
diffusion process near the micro-
electrode. Conunercial nitrogen or
hydrogen derived from a cylinder
of compressed' gas (usually con-
taining less than 0.05 per cent of
oxygen) may be purified by pass-
ing through a Pyrex tube charged
with copper gauze heated to about
450 °C and then through a wash
bottle to saturate it with, water
vapour before passing through the
,, test solution in the polarographic

ceU; the latter procedure minimises any change in volume of the test solution due
to evaporation. It is more convenient, however, to use pure nitrogen (oxygen free)
which can be purchased in cylinders. ' ’

The influence of temperature has already been discussed. The electrolytic cell

Fig. XVI, 5

643

XVI,5 QUANTITATIVE INORGANIC ANALYSIS

should be immersed in a thermostat bath maintained at ±0.2 "C; for many
purposes a temperature variation of ±0.5 ‘'C is permissible. A temperature of
25 “C is usually employed.

As a precautionary measure to prevent the appearance of maxima, sullicient
gelatin to give a final concentration ofO.005 per cent should be added. The gelatin
should preferably be prepared fresh each day; bacterial action usually appears
after a few day.s. Other maximum suppre.ssors (e.g„ Triton X-100 and methyl
cellulose) are sometimes used.

Two or more electro-active ions may be determined successively if their half-
wave potentials ditfer by at least 0.4 volt for univalent ion.s and 0.2 volt for
bivalent ions provided that the ions are present in approximately equal
concentrations. If the concentratioms differ considerably, the difference between
the half-wave potentials must be corre,spondingly larger. If the waves of two ions
overlap or interfere, various experimental device.s may be employed. The half-
wave potential of one of the ions may be displaced to more negative potentials by
the use of suitable complexing agents which are incorporated in the supporting
electrolyte; for example, Cu'* ions may be complexed by the addition of
potassium cyanide. Sometimes one ion may be removed by precipitation (e.g,
with lead and zinc, the lead can be rendered harmless by precipitation as
sulphate; the lead sulphate formed need not be removed by filtration); the
possibilities of adsorption or co-prccipitation of part of the other ions must,
however, be borne in mind. HIcctrolytic .separations are very useful and have the
advantage over chemical separations of eliminating the necessity of adding large
amounts of reagents whose presence during the sub.scqucnl steps of the procedure
may be undesirable, or which may contain traces of the substance being
determined and so give ri.se to an excessively high blank. Electrolytic methods
often provide cleaner separations and arc more rapid. Electrolysis of an acid
solution of the mixture between a large stirred mercury cathode and a platinum
anode may be used for the removal of many clement.s. including Fe, Cu, Ni, Mn,
and Cr. which arc reduced to the metallic state, from others such as .Al. Ti, V, W,
U, the alkaline-earth and alkali rneial.s. Such electrolyses have been used as
procedures in the polarographic determination of vanadium and of aluminium in
steels and alloys. It is better to apply electrolysis with controlled potential at the
mercury cathode whereby much liner separations can be made. Thus to
determine nickel and zinc in a ‘pure’ copper .salt, the latter can be dissolved in an
ammoniacal ammonium chloride solution and electrolysed at about - 0.7 volt vs.
S.C.E. : copper is reduced to metal while nickel and zJnc arc unalfected and can be
determined polarograpliieally in the residual solution. .As little as 0,00001 per
cent nickel or zinc can be determined in this way.

XVl, 5. EVALU.VTIO.N OF QUANTITATIVE RESUL’IiS. Three methods
which have been widely used in practice will be described.

■-^^Vavc height concentration plots. Solutions of several dilTcrcnt concen-
trations of the ion under investigation are prcptircd, the composition of the
supporting electrolyte and the amount of maximum suppressor added being
the same for the comparison standards and for the unknown. The heights of the
waves obtained arc measured in any convenient manner and plotted asa function
of the concentration. The polarogram of the unknown is produced exactly as the
standards, and the concentration is read from the graph. The method is strictly
empirical, and no assumptions, except correspondence with the conditions of the

644

VOLTAMMETRY XVI, 6

calibration, are made. The wave height need not be a linear function of the
concentration, although this is frequently the case. For results of the highest
precision the unknown should be bracketed by standard solutions run

consecutively. ■

Internal standard (pilot ion) method. The relative diffusion currents of ions
in the same supporting electrolyte are independent of the characteristics of the
capillary electrode and, to a close approximation, of the temperature. Hence
upon determining the relative wave heights with the unknown ion and with some
standard or ‘pilot’ ion added to the solution in known amount, and comparing
these with the ratio for known amounts of the same two ions, previously
determined, the concentration of the unknown ion may be computed. This
procedure has limited application, primarily because only a small number of ions
are available to act as pilot or reference ions. The main requirement for such an
ion is that its half-wave potential should differ by at least 0.2 volt from the
unknown or any other ion in the solution with which it might interfere. When a
single unknown is present, this condition can usually be satisfied, but in complex
mixtures there is seldom sufficient difference between the half-wave potentials to
introduce additional waves.

Method of standard addition. The polarogram of the unknown solution is
first recorded, after which a known volume of a standard solution of the sanae ion
is added to the cell and a second polarogram is taken. From the magnitude of the
heights of the two waves, the known concentration of ion added, and the volume
of the solution after the addition, the concentration of the unknown may be
readily calculated as follows. If is the observed diffusion current ( = wave
height) of the unknown solution of volume V cm^ and of concentration C^, and 1 2
is the observed diffusion current after v cm^ of a standard solution - of
concentration C, have been added, then according to the Ilkovic equation we
have: ' . '

h = kC,

and l 2 = k{VC, + vCJiV+v)

Thus k = l 2 {V + v)/{VC^ + vC,)

whence

IjvC,

The accuracy of the method depends upon the precision with which the two
volumes of solution and the corresponding diffusion currents are measured. The
material added should be contained in a medium of the same composition as the
supporting electrolyte, so that the latter is not altered by the addition. The
assumption is made that the wave height is a linear function of the concentration
in the range of concentration employed. The best results would appear to be
obtained when the wave height is about doubled by the addition of the known
amount of standard solution.

XVI, 6. MEASUREMENT OF WAVE HEIGHTS. With a well-defined
polarographic wave where the limiting current plateau is parallel to the residual
current curve, the measurement of the diffusion current is relatively simple In the
exact procedure, illustrated in Fig. XVI, 6, the actual residual current curve is
determined separately with the supporting electrolyte alone; by subtracting the

645

XVI, 6 QUANTITATIVE INORGANIC ANALYSIS

residual current from the value of the current at the dilTusion-currenl plateau
(both measured at the same applied voltage), the dilTusion current is obtained. It
may be noted that when employing polarograms produced with a pen recorder, a
line is drawn through the mid points of the recorder o.scillations. For subsequent
electroactive substances, the dilTusion current would be evaluated by subtracting
both the residual current and all preceding diflusion currents.

It is simpler, though less c.xact, to apply thee.xtrapolation method. The part of
the residual current curve.preceding the initial rise of the wave Is extrapolated: a
line parallel to it is drawn through the diffusion current plateau as shown in Fig.
XVI, 7. For succeeding waves, the diffosion current plateau of the preceding wave
is used as a pseudo -residual current curve.

When the wave is ,v//i,>/i//v distorted the following graphical method may be
used. The distortion of the wave is shown somewhat e.vaggerated in Fig. XVI, S.
Draw the lines AB and CD perpendicularly to the abscissa axis, and divide these
at F and G so that AF= FB and CG= GD. The intersection of FG with the

646

. VOLTAMMETRY XVI, 7

wave, i.e., at £i/ 2 > gives the position of the half-wave potential. Draw the vertical
line HK through the point £ 1 / 2 : HK is the wave height or diffusion current. This
method gives wave heights which are obviously too low when the polarogram is
very much distorted. In practice, extensive distortion of polarograms can often
be averted by application of a ‘compensating’ or ‘counter’ current: see Section
XVI, 8.

XVI, 7. MANUAL NON-RECORDING POLAROGRAPHS. The essen-
tial requirements for measuring polarographic current-rvoltage curves are:

1. a means of applying a variable and known d.c. voltage ranging from 0 to 2 or 3

volts to the cell, and ,

2. a method for measuring the resulting current, which is usually less than about
50 microamperes.

The applied voltage should be known to about 1 millivolt or better, and, the
current measuring device should have a sensitivity of at least 0.01 microampere.
The simple circuit shown in Fig. XVI, 9, may be used. The voltage applied to the
cell is controlled by the potential divider (50- or 100-ohm radio-type
potentiometer), which is powered by two 1.5-volt dry cells. The current is
determined by measuring with the potentiometer the IR drop across the 10000-
ohm precision fixed resistance R 2 . By reversing the double-pole double-throw
switch, the voltage applied to the cell may be measured with the same
potentiometer: any potentiometer with a precision of +0.1 millivolt may be used.

Fig. XVI, 8

Potentiometer

Fig. XVI, 9

The range of the potentiometer should be from 0 to 3 volts ; if the range is 1.6 volts
it can be extended to 0-3.2 volts by standardising against the Weston standard '
cell with the potentiometer set at one-half the voltage of the standard cell and
then miultiplying the observed readings by 2. The galvanometer used as a null-
point detector in the potentiometer circuit should have a period three to four
times longer than the drop time. The characteristics of the galvanometer should
be approximately as follows; sensitivity (critically damped) 0.005-0.01

^mri'nnn damped period 3-5 seconds, internal resistance

about 1000 ohms, and critical damping resistance about 7000 ohms; a resistance

647 •

XVI, 8 QUANTITATIVE INORGANIC ANALYSIS

of 500-1000 ohms across the terminals will increase the period to the desirable
value.

The current-voltage curve is obtained by increasing the applied voltage
stepwise and measuring it with the potentiometer. The current is computed from
Ohm’s Law / - £/Kj. If £2 = 10000 ohms, then each millivolt potential across
Rj corresponds to* lO'VlO'* amp or 0.1 microamp. To e,-<pedite the
me:isureraents, the entire voltage range of interest should first be e.xplored using
fairly large voltage increments, e.g., 0.2 volt, and measuring the current at each
value of the applied voltage. Tlie points should be plotted as the measurements
are made so that the general character of the current-voltage curve becomes
apparent; appropriate additional points required to define the rapidly changing
parts of the curve may then be taken using small voltage increments, say, of 0.01
volt.

Manual polarographs can be purchased, and are useful for mastering the basic
techniques and for understanding the principle.s involved, but they are necessarily
somewhat slow to operate and consequently are rarely used for routine analytical
procedures.

XVI. 8. COMMERCIAL POLAROGRAPHS. In commercial poiaro-
graph.s, provision is made for carrying out the voltage scan automatically by
continuous, steady adjustment of the potentiometer, and at the s;ime time
plotting ihee.m.f. values and the corresponding currents on a chart recorder: the
polarograin is thus obtained immediately and very rapidly. Many polarographs
incorporate a •counter current' control which applies a small opposing current
and can be adjusted to compensate for the residual current; this leads to better
defined polarograms. Although there arc a number of instruments available
which can be u.scd for d.c. polarography. most modern instruments are also
equipped to carry out operations which fall in the realm of a.c. polarography and
will therefore be referred to in that connection. Recent developments in d.c.
polarography are discu.s.scd in Ref. 10.

A useful feature of some commercial polarogmphs is the facility of plotting
derivative polarograms, i.c.. curves obtained by plotting ilI/dE against E. Such
curves show a peak at the half-wave potential, and by ine.isuring the height of the

|)eak It is possible to obtain quantitative
data on the reducible substance: the
height of the {x*ak is proportional to
the concentration of the ion being
di.scharged.

A typical conventional polarograin
for 0.003 .t/-cadmium sulphate in 13/-
potassiuin chloride in the presence of
0.001 per cent gelatin, and the cor-
responding derivative curve is shown in
Fig. XVI. 10. is the ma.xinuira

current recorded on the galvanometer in
the derivative circuit.

The derivative circuit is used for
measuring half-wave potentials closer
S.CE Volts than 0.15 volts, since with the direct
wave, the residual current of the second

648

VOLTAMMETRY XVI, 8

step is always affected by the ion previously discharged. Also in those cases
where the element with the lower half-wave potential is present in much higher
concentration, e.g., in the analysis of copper for cadmiuin content, the analysis is
almost impossible without previous chemical separation. When the norrnal
polarographic waves are differentiated and the rate of change of current with
voltage is recorded (i.e., derivative polarograms) the series of peaks obtained in
positions approximating to the half-wave potentials enable one to identify the
individual elements. Fig. XVI, 11, illustrates the polarogram (A) obtained with
copper and cadmium ions in the ratio of 40 to 1, and the corresponding derivative
polarogram (B) : the two peaks are clearly visible.

-0’3 -0'4 -0-5 -0-6 -0-7 -0'8 -0-9 -0-2 -0-3 -0-4 -O’S -0-6 -0-7 -0'8 -0'9

(A) Eqmevs. Mercury pool anode, volts (B) Edme ''S- Mercury pool anode, volts

Fig. XVI, 11

Many modern polarographs provide for potentiostatic control of the dropping
electrode potential; this is particularly valuable when solutions of high resistance
are involved, as for example when non-aqueous solvents (or mixtures of water
and organic solvents) are used. A high resistance in the polarographic cell leads to
a large Ohm’s law voltage drop (/ x R) across the cell, which not only influences
the measured electrode potential but may also distort the polarogram; in extreme
cases with some non-aqueous solvents, a straightforward I vs. E plot may appear
to be virtually a straight line, and only after correction for the ohmic voltage drop
is a normal polarogram obtained.

Potentiostatic control requires the introduction of a third electrode (a counter
electrode) into the polarographic cell thus leading to an arrangement such as that
depicted in Fig. XIII, 2 (Section XIII, 2). The ‘cell’ in that diagram is now the
polarographic cell, the negative electrode is the dropping mercury electrode,
S.C.E. is the reference electrode (saturated calomel or other suitable electrode),
and the positive electrode is the counter electrode.

The reference electrode must be sited as close as possible to the D.M.E. so that
the resistance of the solution between the two electrodes is reduced to a minimum,
and the potentiostat then maintains the e.m.f. of the.D.M.E.-reference electrode
combination at the correct value. This arrangement has the further advantage,
that no sensible current is passed through the reference electrode and hence there

649

XVI, 9 QUANTITATIVE INORGANIC ANALYSIS

is no possibility of polarisation of the latter with conseciuent variation in
potential

At least one manufacturer (Metrohm Ltd, Herisau, Switzerland) has developed
a system of rapid polarography in which the flow of mercury througli the dropping
electrode is increased mechanically, and the voltage sweep rate increased in
proportion, so that the polarogram is recorded in about one minute, i.e., in about
one-tenth of the time normally required. Apart from the saving in time, this
procedure has the result that much less damping is required in the recording
circuit and in consequence, neighbouring waves are more clearly resolved.

Another feature available in some modern polarographs is the facility of linear
sweep polarography. In the normal polarographic procedure de.scribed in the
preceding sections, the total voltage sweep of the polarogram is spread out over a
succession of several mercury drops, and the rate of change of potential may be of
the order of 10 mV per second. In the linear sweep method the rate of potential
change is greatly increased, up to say 500 mV per second, with the result that the
polarogram is .scanned within the lifetime of a single drop, and in fact, to minimise
the effect of changing drop area, which is greatest when the drop is small and least
near the end of the lifetime of the drop, the potential sweep is accomplished in a
short interval just prior to the fall of the drop. This necessitates the use of a fast
recording device, and in practice an o.scilloscope is usually employed.

Owing to the rapid rise in potential of the D.M.E. the resulting current also
incretises rapidly, and initially it is not diffusion controlled. However, with the
consetiuont discliarge of the reducible ions near to the electrode surface, diffusion
from the bulk solution sets m and the curiciU then falls to the diffusion current
level: in other word.s a peaked polarogram is obtained, somewhat similar in
character to a derivative polarogram. This leads to better resolution than that
achieved in a normal polarogram, and since the peak current is larger than the
diffusion current, greater sensitivity results: it is claimed that concentrations
down to 5 X 10 may he determined. L'.vamplcsofthcuseofthis technique are
discussed in Ref, 16.

XVT, 9. A.NCILLARY EQUIPMENT FOR POLAROGRAPHY. Mer-
cury. Doubly distilled mercury i.s u.sually recommended for polarographic
work. The rc-distillcd mercury of commerce is generally satisfactory for most
determinations: it should be filtered througli a liltcr-papcr cone with a small pin-
hole in the tip (or through a sintercd-glass funnel) before u.se in order to remove
any surface oxides or dust.

Used mercury should be washed with water, thoroughly agitated for about 12
hours in contact with 10 per cent nitric acid (a filter tlask, arranged to admit air
through the bottom of the mercury and connected to a water pump, is
satisfactory), then thoroughly washed with distilled water, dried with filter paper,
and re-distilled under reduceii pressure.

Cautio.n. Mercury vapour is a cuimflative poison. .All vessels containing
mercury should be stoppered. Any .spilled mercury .sboul3 be immediately
collected and placed in a flask containing water, and the bench (floor) dusted with
powdered sulphur. Employ a tray under all vessels containing mercury and for all
operations involving the transfer of mercury.

Dropping mercury electrode a.ssembly. The ti-sscmbly consists of a mercury
reservoir (c.g., a lOO-cnv* levelling bulb), a connecting-tube between the reservoir
and the capillary tube, and a small glass electrolysis cell in which the unknown

650

VOLTAMMETRY XVI, 9

solution is placed. 'A simple arrangement is
shown in Fig. XVI, 12. The heavy-walled rubber
tubing, 80-100cm long, should be sulphur-free.
Neoprene tubing is generally ernployed; the
inside surface should be steamed out for 30
minutes before use, followed by drying with air
filtered through a cotton filter-plug. The elec-
trolysis cell shown in the figure is the original
type devised by Heyrovsky. Electrical con-
nection to the mercury in the reservoir is affected
by a platinum wire sealed into the end of a soft-
glass tube, which is partly filled with mercury
and held in place by the stopper of the reservoir.

The effective capillary tube has a length of
5-10 cm and a bore diameter of about 0.05 mm
(range 0.04-0.07 ihm); the outside diameter is
usually about 6-7 mm: the delivery tip is cut
accurately horizontal. Suitable capillary tubes
may be purchased from any manufacturer of
commercial polarographs. At a given pressure
the drop time (which is the time that elapses
between the fall of two successive drops) is directly proportional to the length of
the capillary, but inversely to the third power of its internal radius ; it is also
inversely proportional to the pressure on the drop. A capillary suitable for use in
polarography should have a length and bore such that the application of a
pressure of about 50 cm of mercury will cause a drop weighing 6^10 mg to fall
every 3-6 seconds when the top of the capillary is immersed in distilled water. The
dropping mercury electrode must be mounted so that it is within + 5° of the
vertical ; deviation from this angle produces erratic dropping.

With careful treatment, a capillary should remain serviceable for many
months. It is absolutely essential that no solid matter of any kind should be
allowed to reach the inside of the capillary. The electrode must never be allowed
to stand in a solution when the mercury is not flowing.

The following procedure is recommended. The sample solution is deaerated,
then, with the tip of the capillary in the air, the mercury pressure is raised at least
10 cm above the previously found equilibrium height, the capillary is inserted into
the cell, and the mercury level is finally adjusted to the desired value. After the
completion of the measurements the capillary is withdrawn from the cell and
washed thoroughly with a stream of water from a wash bottle while the mercury is
still issuing from the tip and is being collected in a micro beaker. The mercury
reservoir is then lowered until the mercury flow just ceases (not further) and the
electrode is allowed to stand in the air. It is good technique, at the beginning of
each period of use, to immerse the capillary for ca. 1 min. in 1: 1-nitric acid while
mereury is flowing, then wash it well with distilled water; a further precaution
is to allow the mercury drops to form in distilled water for about 15 minutes.

If the capillary becomes partly or completely blocked it is sometimes possible
to clear it by carefully drawing strong nitric acid through it until the foreign
matter has been completely dissolved, followed by distilled water to remove all
traces of acid; the capillary is finally dried by drawing a stream of warm air
(filtered through a cotton-wool plug) through it. .

651

XVr,9 QUANTITATIVE INORGANIC ANALYSIS

For reproducible results with the same dropping mercury electrode, it is
important that the height of the mercury in the reservoir above the capillary tip
should be constant, i.e., the same pressure on the dropping mercury tip be
maintained : the small quantity of mercury passing through the capillary does not
appreciably alTcct the volume in thereservoir. Fhestand supporting the electrode
should permit the rapid immersion and remov.'il of the capillary from tlie solution
in the polarographic cell, particularly when the latter is in position in a
thermostat. Suitable stands are available from the manufacturers of
polarographs.

Polarographic cells. Numerous types of polarographic cells have been
described and various forms are available commercially; the choice may well be
dictated by the electrode stand in use.

The original Heyrovsky cell is depicted in Fig. XVT, 12. and can be readily
constructed in the laboratory from a conical llask. The H-type cell devised by
Lingane and Laitincn and shown in Fig. XVI. 13, will be found satisfactory for
most purposes; a particular feature is the built-in reference electrode. Usually a

winple

Fig. XVI, 13

saturated calomel electrode is employed, but if the presence of chloride ion is
harmful a mercuryllj sulphate electrode (llg,Hg,SO.t in potassium sulphate
solution: potcntialcu. t-O.dO volt vs. S.C.E.imay be used, It is usually designed to
contain 10~.50 enf* of the sample solution in the left-hand compartment, but it
can be constructed to accommodate a smaller volume down to 1-2 cm-*. To avoid
polarisation of tlic reference electrode the latter should be made of tubing at least
20 mm in diameter, but the dimensions of the solution compartment c;ui be
varied over wide limits, 'fhe compartments are separated by a cross-member
filled with a 4 per cent agar-s;iiuraicd potassium chloride gel, which is held in
position by a medium-porosity sintered Pyre.v glass disc (diameter at least 10 mm)
placed as near the solution compartment as possible in order to facilitate
deaeration of the test solution. By clamping the cell so that the cross-member is
Vertical, the molten agar gel is pipetted into the cross-member and the cell is
allowed to stand undisturbed until the gel has solidified.

In use, the solution compartment (either dried by aspiration of air through it or

652

VOLTAMMETRY XVI, 10

rinsed with several portions of the test solution) is charged with at least enough
test solution to cover the entire sintered-glass disc. Dissolved air is removed by
bubbling pure nitrogen through the solution via the side arm : by means of a two-
way tap in the gas stream, the gas is then diverted over the surface of the solution.
Measurements should not be attempted while gas is bubbling through the
solution, for the stirring causes high and erratic currents. Finally, the dropping
electrode is inserted through, another hole in the stopper (which should be large
enough for ease of insertion and removal of the capillary) and the measurements
are made. When the H-cell is not in use the left-hand compartment should be kept
filled either with water or with saturated potassium chloride solution (or other
electrolyte appropriate to the reference electrode being used) to prevent the agar
plug from drying out.

Maximum suppressors. Gelatin is widely used as a maximum suppressor in
spite of the fact that its aqueous solution deteriorates fairly rapidly, and must
therefore be prepared afresh every few days as needed. Usually a 0.2 per cent stock
solution is prepared as follows. Allow 0.2 g of pure powdered gelatin (the grade
sold for bacteriological work is very satisfactory) to stand in 100 cm^ of boiled-
out distilled water for about 30 minutes with occasional swirling: warm the flask
containing the mixture to about 70 “C on a water bath for about 15 minutes or
until all the solid has dissolved. The solution must not be boiled or heated with a
free flame. Stopper the flask firmly. This solution does not usually keep for more
than about 48 hours. Its stability may be increased to a few days by adding a few
drops of sulphur-free toluene or a small crystal of thymol, but, the addition is
rarely worth while and is not recommended.

A gelatin concentration of 0.005 percent, which corresponds to 0.25 cm^ of the
stock 0.2 per cent solution in each 10 cm^ of the solution being analysed, usually
suffices to eliminate maxima. Higher concentrations (certainly not above 0.01 per
cent) should not be used, since these will distort the wave form and decrease the
diffusion current markedly.

Triton X-100, like gelatin, suppresses both positive and negative maxima, but,
unlike gelatin, its aqueous solution is stable. A stock 0.2 per cent solution is
prepared by shaking 0.20 g of Triton X-100 thoroughly with 100 cm^ of water.
About 0.1 cm^ of this solution should be added to each 10 cm^ of the sample
solution to give a Triton X-100 concentration of 0.002 per cent. ,

XVI, 10. DETERMINATION OF THE HALF-WAVE POTENTIAL OF
THE CADMIUM ION IN M-POTASSIUM CHLORIDE SOLUTION. The
following experiments (Sections XVI, 10-XVI, 12), which can well be performed
with a manual polarograph, serve to illustrate the general procedure to be
followed in d.c. polarography.

Follow the operating instructions for the particular apparatus in use. Make sure
that the reservoir of the dropping electrode contains an adequate supply of
mercury, and that mercury drops freely from the capillary when the tip is
immersed in distilled water whilst the reservoir is raised to near the maximum
height of the stand: allow the mercury to drop for 5-10 minutes. Replace the
beaker of water by one containing M-potassium chloride solution and adjust
the rate of dropping by varying the height of the mercury reservoir until the
dropping rate is 20-24 per minute: then clamp the mercury reservoir in position.

When the measurements have been completed {vide infra), rinse the capillary
well with a stream of distilled water from a wash bottle and then dry by blotting

653

XVI, 10 QUANTITATIVE INORGANIC ANALYSIS

with filter paper. Insert the capillary through an inverted cone of quantitative
filter paper and clamp vertically over a small beaker. Lower the levelling bulb
until the mercury drops just cease to flow.

Pipette 10 cm^ of a cadmium sulphate solution (l.OgCd^*^ dm"-') into a 100-
cm-' measuring flask, add 2.5 cm^ of 0.2 per cent gelatin solution, 50 cm' of
2.U-potassiuni chloride solution and dilute to the mark. The icsulting solution
(/I) will contain 0.100 g Cd'*^ dm"-' in a base .solution (supporting electrolyte)
of .l/-potassium chloride with 0.005 per cent gelatin solution as suppressor.

McasuremeiiLs. Place 5.0 cm' of the solution A in a polarographic cell
equipped with an external reference electrode (.saturated calomel electrode). Pass
pure nitrogen through the solution at a rate of about 2 bubbles per second for lO-
15 minutes in order to remote dissolved oxygen. Raise the mercury reservoir to
the previou-sly determined height and insert the capillary into the cell so that the
capillary tip is immersed in the solution. Connect the S.C.E. to the positive
terminal and the mercury in the reservoir to the negative terminal of the
polarograph. After about 15 minutes slop the p;issage of inert gas through the
solution: the electrical measurements may now be commenced.

Carry out a preliminary test to ascertain the optimum position of the
galvanometer shunt. Slowly turn the applied potential dial and depress the
tapping key (or equivalent switch) at intervals. It will be found that at a certain
point the current (as indicated by the deflection of the galvanometer spot) will
increase rapidly. Decrease llie galvanometer sensitivity by means of the shunt
switch until the spot remains on the scale even at ma.ximum applied potential; the
latter should not. of course, e.xcced the decomposition potential of the supporting
electrolyte.

Now, with tile desired .sensitivity in circuit and commencing from zero. incrc;be
the applied potential in suitable steps (.say, in 0.05 v oh) and read the galvanometer
deflection for each value of the applied voltage. When the deposition potential of
the cadmium ion is reached, tlie galvanometer deflection (i.e., current) increases
rapidly, and smaller increments of the applied voltage (s;iy. in 0,01 volt steps) are
then advi.sable until the rate of change decrc.ises considerably. It should be noted
that the inaximtiin deflection of the spot must be recorded. Plot the applied
voltage (abscissae) against current (ordinates) .IS represented by the galvanometer
deflections. (The actual current flowing at each value of the applied voltage may-
be easily computed, if desired, from the known sensitiv iiy of the galvanometer, as

determined by the manufacturers, and
the position of the sensitivity switch.)
The graph .should have the form shown
in Fig. XVI, 4. Determine the half-wave
potential from the current-voltage
curve as described in Section XVI, 5;
the value in AZ-potussium chloride
should be about ~ 0.60 volts is, the
S.C.E Alternatively, measure the
maximum height of the dilTusion wave
after correction has been made for the
residual current; this is the dilTusion
current Jj, and is proportional to the
, . total concentration of cadmium ions in
Fig- XVI. 14 Los-f- the solution.

6.54

VOLTAMMETRY XVI, 11/12

Measure the height of the diffusion wave I, after correcting for the residual
current at each increment of the applied voltage. Plot the values of log (/^ — /)/ 1 as
abscissae against applied voltage as ordinates (Fig. XVI, 14; strictly speaking, the
values of the applied .voltage are negative). Determine the slope of the graph,
which should be equal to about 0.030, and read off the intercept on the voltage
axis. The latter is the desired half-wave potential of the cadmium ion vs. S;C.E.
.As an additional exercise, the current-voltage curve of the supporting
electrolyte (M-potassium chloride) may be evaluated; this gives the residual
current directly and no extrapolation is required for the determination of 1 and 1^.

XVI, 11. DETERMINATION OF CADMIUM IN SOLUTION. Two
procedures may be employed: (i) that dependent' upon wave height-
concentration plots, and (ii) the method of standard additions. The theory has
been given in Section XVI, 5.

(i) Wave height-concentration plots. Prepare from the stock solution
containing 1.000 g Cd^'*' dm“^ solutions containing respectively 0.1, ,0.05,
0.025, and 0.01 g Cd^"^ dm”^ by transferring to 100-cm^ graduated flasks 10,
5.0, 2.5, and 1.0 cm^ of the stock solution, adding 50 cm^ of 2M-p6tassimn
chloride solution and 2.5 cm^ of 0.2 per cent gelatin solution to each flask and
then diluting to the mark with distilled water. Mix 10 cm^ .of the unknown
solution (which may contain, say, about 0.04 g of cadmium dm"^) in a 100-cm^
measuring-flask with 50 cm^ of 2M-potassium chloride solution and 2.5 cm^ of
0.2 per cent gelatin solution and dilute to the mark. Record the polarograms of
the four standard solutions and of the unknown solution following the procedure
described in Section XVI, 10, and determine the wave heights, from each
polarogram. Draw a calibration curve (wave heights as ordinates, concentrations
as abscissae) for the four standard solutions: read off from the curve . the
concentration corresponding to the wave height of the unknown solution.

(ii) Method of standard additions. The polarogram of the unknown solution
will have been determined under (i). A new polarogram must now be recorded
after the addition of a known volume of a standard solution containing the same
ion, care being taken that in the resulting solution the concentrations of the
supporting electrolyte and the suppressor are maintained constant.

Place 10 cm^ of the unknown solution, 5 cm^ of the stock solution (1.000 g
Cd^"^ dm“^), 50 cm^ of 2M-potassium chloride solution, and 2.5 cm^ of 0.2 per
cent gelatin solution in a 100-cm^ graduated flask, and dilute to the mark with
distilled water. Transfer a suitable volume to the polarographic cell in a
thermostat, remove the dissolved air with nitrogen, and record the polarogram in
the usual way. It is important that the galvanometer sensitivity be kept at the
previous value. Determine the new wave height. Calculate the concentration of
the unknown solution with the aid of the formula given in Section XVI, 5.
Compare this value of the concentration with that found by method (i).

XVI, 12. INVESTIGATION OF THE INFLUENCE OF DISSOLVED
OXYGEN. The solubility of oxygen in water at the ordinary laboratory
temperature is about 8 mg (or 2.5x10"“^ mol), per litre. Oxygen gives two
polarographic waves (O 2 — >■ H 2 O 2 — >• H 2 O) which occupy a considerable
voltage range, and their positions depend upon the pH of the solution. Unless the
test solution contains a substance which yields a large wave or waves compared
with which those due to oxygen are negligible, dissolved oxygen will interfere. In

655

XVI, 13/14 QUANTITATIVE INORGANIC ANALYSIS

general, particularly in dilute solution, dissolved oxygen must be removed by
passing pure nitrogen or hydrogen through the solution.

Place some A/-pota.s.sium chloride solution containing 0.005 percent gelatin in
a polarographic cell immersed in a thermostat. Make the usual preliminary
adjustments with regard to .sensitivity control of the galvanometer, observe tlie
current (galvanometer dellection) at increasing values of applied voltage, and plot
the current-applied voltage curve. Now pass oxygen-free nitrogen through the
solution for 10-15 minutes. Plot the polarogram using the .same galvanometer
sensitivity. It will be observed that the two oxygen waves are absent in the new
polarogram (compare Fig. XVI, 5).

XVI, 13 . DETERMIN.VnON OF LEAD AND COPPER IN STEEL In
the application of the polarographic method of analysis to steel a serious
difliculty arises owing to the reduction of iron(I!lJ ions at or near zero potential in
many base electrolytes. One method of surmounting the dilficulty is to reduce
iron(Ill) to iron(ll) with hydra/inium chloride in a hydrochloric acid medium.
The current near zero potential is eliminated, but that due to the reduction of
iron(ll) ions at about - 1.4 volt r.s. S.C.E. still oceur.s. Other metals (including
copper and lead) which are reduced at potentials less negative than this can then
be determined without interference from the iron. Alternatively, the Fe^ ' to Fc" *
reduction step may be sliifted to more negative poicnliaks by complex ion
formation.

The following procedure may be used for the simultaneous determination of
copper and lead in plain carbon steels. Dissolve 5.0 g of the steel, accurately
weighed, in a mixture of 25 cm-* of w ater and 25 ciif* ofconcentraled Itydrochloric
acid: heal gently to minimise the loss of acid. Aiid a few drops of saturated
potassium chlorate .solution to dissolve carbides, etc., and hoi! the mixture until
the .solution is clear. Cool and dilute to 50 cm^ with water in a graduated tla,sk.
Pipette 2.00 cm-* of this solution into a polarographic cell and add: 1,0 cm^ of 20
per cent hydrazinium chloride solution to reduce any iron(Ul) to iron(Ii) state, 1.0
cm ' of 0.2 per cent methyl cellulose to act as a maximum suppressor, and 5.5 cm^
of2.0.\/-.sodium formate .solution to adjust the pH of the solution to that at wiiieh
reduction of Fe(lll) and Cu(Il) ions takes place. Place the celt in a nearly boiling
water bath for 10 minutes in order to complete the reduction. Cool. Analyse the
solution polarographiculiy : u.sc a saturated calomel reference electrode. 'llie first
step in the polarogram is due to the reduction of coppcr(l) ions to the metal and
has a half-wave potential of — 0.25 volt f.v. S.C.Ii. Tlie .second step, which is due to
lead, has a half- wave potential of - 0.45 volt v.s. S.C.E. Carry out a calibration by
adding known amounts of copper and lead to a solution of steel of low copper and
le;id content, and determine the increase in wave heights due to the additions.

Calculate the percentage of copper and of lead in the sample of steel.

z\lteriiating current polarography

XVl, 14. THE NATURE OF a.c. METHODS. The u.se of alternating
current in polarographic measurements has developed in two distinct ways: (u)
by replacement of the direct current used in d.c. polarography by an alternating
current; (h) by the introduction of an a.c. voltage into a polarographic circuit
operating with direct current. Mcihod.s in which alternating current only is
employed arc referred to under the heading of Oscillographic Polarography, and

656

VOLTAMMETRY XVT, 15

the term a.c. Polarography is restricted to the combined use of a.c. and d.c. : it is in
this field that the greatest advances in polarography have occurred in recent years
{Ref.5). .

XVI, 15. SIMPLE a-c. POLAROGRAPHY. The fundamental features of a
simple a.c. polarograph (Ref. 6) can be indicated by considering the modifications
introduced into the basic circuit shown in Fig. XVI, 1 (Section XVI, 2). The
galvanometer G and voltmeter V were removed, and resistance S was joined to
the dropping mercury electrode through the secondary winding of a variable
step-down transformer, so that when the transformer was switched on, the, 50
Hertz mains voltage produced a resultant a.c. voltage of 1-100 millivolts which
was superimposed on the d.c. current flowing in the polarograph. The measuring
circuit, which was joined across S, consisted of a condenser (to suppress the d.c.
current), an amplifier A, and a valve voltmeter B as in Fig. XVI, 15; the voltage
reading on B coupled with the known value of the resistance S; allowed the value
of the a.c. current to be calculated.

Fig. XVI, 15

Fig. XVI, 16

If the values of the a.c. current are plotted against the potential applied by the
potentiometer, a series of peaks are obtained as illustrated in Fig. XVI, 16 (a); the
normal d.c. polarogram of the same solution is also shown (curve b).

The a.c. curve is seen to be similar in character to a derivative polarogram
(F igs. XVI, 10; XVI, 1 1, B : Section XVI, 8) but must not be confused with this type
01 curve. Each peak in the a.c. curve corresponds to a step in the normal

657

XVI, 16/17 QUANTITATIVE INORGANIC ANALYSIS

polarographic record. The voltage of the peak is the same a.s' that of the midpoint
of the step, and the heiglii of a peak above the base line is proportional to the
concentration of the dcpolariser, and thus corresponds to the step height. It will
be apparent that with closely separated svave.s. measurements arc much more
readily made from the a.c. polarogram than from the d.c. polarogram, and it is
considered that peaks separated by 40 mV can be resolved as compared with the
separation of 200 mV required in d.c. polarography: the limit of sensitivity
(10" ^.\f) is, however, not greatly dilTerent from that achieved in d.c.
polarography, and this is related to the fact that the residual current is rather
large. This arises because the condenser current is relatively large as compared
with thedin'usion or faradaic current.

XVI, 16. SQUARE-WAVE POLAROGRAPHY. Barker and Jenkins (Ref. 8)
attempted to solve the problem arising from the large condenser current by
replacing the normal a.c. sine wave current by a .square wave current, which
means that for a large part of each half cycle, the applied a.c. voltage is constant
and the as.sociaied charging and faradaic currents rise rapidly to their ma,ximum
values. Subsequently each of these currents dccay.s during the half cycle, but
whereas the charging current decays virtually to zero, the faradaic current
decreases slowly, and hence if the current is measured near the end of the half
cycle (over the last 2 x 10 ‘ * sec), the value recorded should be the faradaic current
free from the effect of the charging current. This device for minimising the
intluence of the charging current is termed 'Fast pohirography or strobe
polarography. In fact, however, complications arise because the mercury drop is
still increasing m area during the measurement period, and it was found that this
could be compensated by modifying the completely square wave form into one
with a slight downward slope on the upper edge and a corresponding upward
slope on the lower edge. It was also found necessary to keep the resistance of the
cell low: this could be achieved by increasing the concentration of thesupporting
electrolyte, which, however, needed to be very pure, otherwise Iracx impurities
gave rise to complications. To avoid po.ssib!e interaction arising from mains
frequency harmonics, tlic square wave current was generated at a frequency of
225 Hz.

Elaborate electronic techniques were required to achieve the requisite wave
fonn and frequency and the necessary constant periodiciiy of current
measurement, and although concentrations a,s low as 5xI0~’,Vf could be
determined m certain cases, problems were encountered in connection with the
so-called ‘capillary response' (Ref. I5j. As each mercury drop falls from the
capillary a little aqueous solution may be drawn into the tip of the capillary: the
amount varies with each drop. The cipacity current of this trapped liquid is found
to be significant e\ cn at the end of the half cycle of the square vv;i ve voltage sweep
thus upsetting the premise upon which the procedure is based. The procedure was
accordingly modified by Barker and Gardner (Ref. 9) to give rise to what is
termed pulse polarography,

XVl, 17. PUI.se polarography. Barker and Gardner argued that
given sullicicnt time, the disturbing clfects due to the ‘capillary rcspon.se' would
disappear, and if therefore the frequency of the square wav e current were reduced,
thus giving a longer period between measurements, the required result would be
achieved. It was discovered, however, that the requisite reduction in frequency

658

VOLTAMMETRY XVI, 17

was so great as to be impracticable; it would need to be reduced from the 225 Hz
of the square wave polar ograph to 10 Hz. . •

The solution of this problem was found by replacing the square wave current
by a series of potential pulses; one pulse of approximately 0.05 second duration
being applied during the growth of a mercury drop, and at a fixed point near the,
end of the life of the drop. Two different procedures may, however, be employed:
(a) pulses of increasing amplitude, may be superimposed upon a constant d.c.
potential, or [b) pulses of constant amplitude may be applied to a steadily
increasing d.c. potential. • '

In method (a) the onset of the pulse is marked by a sudden rise in the total
current passing: this is largely, due to the condenser (charging) current which,
however, soon decays to zero. The.faradaic current also decays, but only to the
level of the diffusion current, and if the current measurement is made in the last
stages of the pulse (in the final 33 ms of its duration), it gives the faradaic current
alone. The situation is thus similar to that encountered in square wave,
polarography, but the time scale is long enough, for the disturbing effects of the
‘capillary response’ to be removed. The resulting polarogram is similar to a
conventional d.c. polarogram except that the characteristic saw-tooth pattern of
the latter is replaced by a stepped curve.

In method {b) the applied potential varies with time as shown in Fig. XVI, 17.
The current is read near the end of each pulse, i.e., at points corresponding to C.'
When the faradaic current is small the pulse current will also be small, but as a
normal d.c. polarographic wave sets in and the faradaic current rises, so too does
the pulse current, and it will attain a maximum value at the half wave potential for
the system under investigation. Consequently, the pulse current-d.c. voltage plot
will be a peaked curve similar to a derivative polarogram of conventional d.c.
polarography, and will in fact closely resemble a square wave polarogram.

For this reason, procedure (h), i.e., the application of pulses of constant
amplitude to a steadily increasing d.c. voltage, is often referred to as differential
pulse polarography. The alternative procedure (a) is consequently referred to as
normal pulse polarography or as integral pulse polarography.

, J!” ^PP^''3tus produced by one supplier (Princeton Applied Research, Model
70 Electrochemistry System and also Model 174 Polarographic Analyser), the
current is measured twice during the lifetime of each mercury drop: once just
elore the application of the pulse (points corresponding to C' in Fig. XVI, 17) as

659

XVr, 18 QUANTITATIVE INORGANIC ANALYSIS

well as at the usual point C near the end of the pulse. The current at C' is that
which would be observed in normal d.c, polarography, its value is stored in the
instrument. The onset of the pulse is then marked by a sudden rise in current,
which as in method (u) soon settles down as the conden.ser current decays, and
near the end of the pulse (point C). the current is again read. This value is then
compared with that stored in the instrument (the value for point C), and the
ditrcrence between them is amplified and recorded. Clearly, if the measurements
of current are made on the residual current curve, or on a plateau of the d.e.
polarograni, thedillerence in current between points C and C will be small, butif
the measurements are made on a polarographic wave, an appreciable current will
be recorded, and it will of course reach a maximum value when the applied d.c.
potential is equal to the liulf wave potential. The dilTcrence current plotted
against the tipplied d.c. potential will therefore be a {leaked curve with the height
of the peak proportional to the concentration of the reducible substance in the
solution, just as with a derivative potarogram. DilTcrenlial pulse polarography is
a very satisfactory method for the determination ofmany substances at thep.p.m.
level.

An important feature of both square wave and pulse poiarogra{rhy is the
sampling of the current at definite points in the lifetime of the mercury drop, and it
is essential to establish an e.xact timing procedure. Various methods have been
adopted to achieve the desired result: these include mechanical tapping of the
capillary to dislodge the mercury drop at precise time intervals, or alternatively
the natural drop time of the capillary is utilised, with the fall of the mercury drop
serving to actuate tlie timing circuits which control the mea.surement of the
current.

Oscillographic polarography

With the introduction of a.c methods of ftoJarograpliy, the application of the
cathode ray o,scillosco(>e to polarogr.iphic iiuesligations was an obvious further
step. Developments have followed two mam lines: invcsiigalions with controlled
currents (this was the jrrocedurc adopted initially by Heyrovsky — Ref. 11), and
investigations at controlled potentials (Ref. 12). Results given by the second
procedure most clo,scly resemble those given by other polarographic methods
and will therefore be con.sidered first.

XVI, 18. COiST ROLLED POTE.VHAL .METHODS. Some of the
problems cncoiiiiiercd in simple a.c. and in sqn.irc vvave polarography arc also
found in this technique; in (larticular the existence oflargc charging or condenser
currents, and it was Randles (Ref. 12) who first used the technique of a volt.ige
pulse applied near the end of the lifetime of the mercury drop to overcome this
diOiculty. The voltage pulse covered the complete voltage range required for
the particular polarogram, and for appro.ximatcly two-thirds of its lifetime the
mercury drop was allowed to grow with no voltage applied. Then over the
remaining one-third of the life span, the complete voltage sweep was applied so
that the maximum voltage was reached immciliatcly before the drop fell. When
the drop fell, the applied voltage simultaneously dropyicd back to zero arid
consequently, the applied voltage- time plot was of a s;iw tooth character.

The resultant display on the oscilloscope .screen wa.s .somewhat similar to a

660

VOLTAMMETRY XVI, 19

conventional polarogram showing a maximum (Fig. XVI, 3, Section XVI, 3), and
it was established that there is a linear relationship between the peak current and
the concentration of the depolariser as shown in the Randles-Sevcik equation:

Jp = 2344n"'^ m^/^ t"^3 C

where is the peak current, a the rate of voltage sweep (volts per second), and the
other terms have the same significance as in the Ilkovic equation (Section XVI, 3).

The replacement of the conventional polarogram by a peaked curve leads to
increased sensitivity and better resolution than is achieved in normal d.c.
polarography: the peak current is approximately 4x^/n times the diffusion
current of normal polarography, where n is the number of electrons involved in
the electrode reaction.

Davis and co-workers (Ref. 13) further developed the above observations to
produce a polarograph suitable for analytical purposes. A particular feature of
this work is the development of a dual dropping electrode system. If the two
electrodes, both dropping at the same rate, and with capillaries as near alike as
possible, are placed in separate cells, then it is possible to measure the difference
between the currents flowing in the two cells. If one cell contains the test solution,
and the other contains the pure base electrolyte, then only the current due to the
component to be determined in the test solution will be measured; this procedure
is referred to as subtractive (or differential) polarography.

In comparative polarography a similar procedure is used but with the test
solution in one cell and a reference solution in the second cell: this reference
solution has the same base electrolyte as the test solution plus an accurately
known amount of the substance to be measured. The measured current is now
due to the difference in concentration of the substance to be determined in the test
and reference solutions.

Derivative polarograms can be obtained by using a single dropping electrode
in conjunction with the appropriate electrical circuit of the instrument, and if this
same circuit is employed together with the dual electrode system, then the second
derivative curve is obtained.

XVI, 19. CONTROLLED CURRENT METHODS. In this technique, a
controlled current is passed through the cell and the variation of potential with
time is recorded. In the early investigations of Heyrovsky and Forjet (Ref. 11), a
strearning mercury electrode was employed so as to overcome the effect of the
changing size of the mercury drop in a conventional dropping electrode, but the
latter may be employed if arrangements are made to record the oscilloscope trace
only towards the end of the lifetime of the drop in much the same way as
measurements are made in pulse polarography.

The trace displayed on the oscilloscope may be either E vs. t, dE/dt vs. t or
dEjdt vs. E; for quantitative purposes the last of these is generally found to be the
most satisfactory. Using a base solution free from impurities the resultant trace is
an ellipse, but in the presence of an electro-active substance, indentations appear
on the ellipse giving the result shown in Fig. XVI, 18. The ellipse is symmetrical
about the horizontal axis provided that the electrode reaction is reversible; the
indentations occur at the half wave potential of the electrode reaction concerned,
f he depth of the indentations is a function of the concentration of the substance
giving rise to them, and is most easily measured by employing a dual beam

661

XVI, 20 QUANTITATIVE INORGANIC ANALYSIS

oscilloscope with one beam utilised
to produce the horizontal trace AA'
(Fig. XVI. IS). The height of this
trace can be altered by application of
a voltage derived from a poten-
tiometer and a note is made of the
voltage (£j) required to displace AA'
from the axi.s of the ellipse to the
position shown. If the experiment is
repeated using a solution containing
the dei>olariser at a different con-
centration. a different voltage (£,)
will be required to displace AA' from
the a.\is to the trough of the new
Fie--'^V1, 18 indentation, and proceeding in this

manner, a calibration curve can be constructed in which the voltage is plotted
against the concentration ofdepolari.scr; for further details see Ref. 14.

XVI, 20. INSTUU.MIC.NT A1TON. A numlKr of polarographs are available
commercially which entible various aspects of a.c. polarography to be carried out,
and which usually also make provision for d.c. polarography. Typical
instruments include Druker Eiccirospin Ltd, .Model E 310 Universal Modular
Polarograph (Fig. XVI, I'J); the Princeton Applied Research .Model 174
Polarographic Analyser and the more sophisticated Model 1 70 Electrochemistry

Hr. XVI, 19

662

VOLTAMMETRY XVI, 21

System; the Metrohm ‘Polarecord’ used in conjunction with the a.c. modulator
unit E 393. All such instruments incorporate, or they may be connected to, chart
recorders so that the polarograms are. recorded directly, and in many cases,
connection can be made to an oscilloscope. The Davis Differential. Cathode Ray
Polarograph, Model A1660 -(Fig. XVI, 20), supplied by Shandon. Southern
Instruments Ltd, contains a built-in cathode ray tube: it is designed specifically
for controlled potential oscillographic polarography and is equipped with a dual
dropping mercury electrode system as described in Section XVI, 18.

Fig. XVI, 20

XVI, 21. QUANTITATIVE DETERMINATIONS. Any of the determina-
tions described under d.c. polarography (Sections XVI, 11-13) may be carried out
by a.c. methods: the procedures available will, of course, be governed by the
polarographic facilities available to the analyst. The general procedure to be
fmlowed will be similar to that described for d.c. polarography, with due
observance of the operating instructions for the particular instrument in use. It is
instructive to carry out one of the determinations (copper and lead in steel,
Section XVI, 13, is particularly apt), by a number of the techniques, and to
ascertain the concentration limits which can be determined by each one. Specific
examples of the applications of differential pulse polarography are given in

663

XVI, 22 QUANTITATIVE INORGANIC ANALYSIS

Anodic stripping voltammetry

XVI, 22, B.ISIC PRINCIPLES. If a convcntiaiial d.c. polarographic system
is set up; with an oxygen-free solution containing one or more ions reducible at
the DME, and with the height of the mercury column of the electrode carefully
reduced until dropping has ceased, so that a single mercury drop is left attached
to the capillary, we have a Hanging Mercury Drop Electrode (HMDE) system. If
the potentiometer of the polarograph is then set to ii fixed value which is chosen to
be 0,2“0.4 V more negative than the highest reduction potential encountered
amongst the reducible ions, then electrolysis will occur, deposition of metals will
take place on the HMDE cathode and usually, amalgam formation will take
place. The rate of amalgam formation will be governed by the magnitude of the
current llowing, by the concentrations of the reducible ions, and by the rate at
which the ions travel to the electrode; the latter can be controlled by stirring the
solution. Given sullicient titne, virtually the whole of the reducible ion content of
the solution may be transferred to the mercury cathode, but complete exhaustion
of the solution is not really necessary for the pre.scnt procedure, and in practice,
electrolysis is carried out for a carefully controlled time interval .so that a fraction
(say 10 per cent) of the reducible ions arc iliscliLirged. This operation is often
referred to as a concentration step; the metals beeonte concentrated into the
relatively small volume of the mercury drop.

If the connections to the cell arc now tc\cr.sed, and the potentiometer of the
polarograph set to its lowest range and titen allowed to vary in the norma!
manner (tlie potentiometer siiould be motor driven to ensure a steady rate of
change of potential), then a gradually increasing po.sitivc potential is applied to
the HMDE which is now the anode of the celt. If the current is measured and
plotted against the anodic voltage {a recorder is used}, then initially a gradually
increasing current corresponding to the residual current of conventional
polarography, and due mainly to tlte ground solution, is observed. ,As the anodic
potential approaches the oxidation potential of one of the metals dissolved in the
mercury, then ions of that met.il (i.tss into solution from the amalgam and the
current increases rapidly and attains a maximum value when the anodic potential
has a value approviniaiing to the appropriate oxidation potential. The metal is
said to be stripped from the amalgam, and if the potential were held at the value
corresponding to the maximum current, all of the metal would eventually he
returned to the .solution. In actual fact, however, the potential is not held
stationary, and as the potential sweep continues, the current declines from its
maximum value and settles down to a new approxim.iiely steady value; in other
words, the curve shows a peak. With continuing rise in the anodic potential, fresh
peaks will be producevl in the curve as the oxidation potentials of the dilTercnt
metals contained in the amalgam arc reached; by analogy with polarogram, the
resulting curve is termed a voltamniogram (or stripping voltununogram,).

The peaks are characterised by the peak potential £p, by the peak current (i.e.,
the height of the peak ) and by the breadth /> (i.c.. the voltage span of the peak at the
point where the current is 0.5/^,); tlicse paramctci>i are, however, dependent upon
characteristics of the electrode and upon the rate of the voltage sweep during the
stripping proces.s. The magnitude of the peak current is proportional to the
concentration in the amalgam of the metal being stripped, and is therefore
proportional to its concentration in the original solution.

From the nature of the process described above it has been referred to as

664

VOLTAMMETRY XVI, 23

inverse polarography and also as stripping polarography, but the term anodic
stripping voltammetry is preferred (Ref. 17). It is also possible to reverse the
polarity of the two electrodes of the cell thus leading to the technique of cathodic
stripping voltammetry.

In just the same, way as differential pulse polarography represents a vast
improvement over conventional polarography (see Section XVI, 17), the
application of a pulsed procedure leads to the greatly improved technique of
differential pulsed anodic (cathodic) stripping voltammetry. A particular feature of
this technique is that owing to the much better resolution and greater sensitivity
which is achieved, the concentration of metals in the HMDE can be reduced and
consequently, the time needed for the concentration step can be cut considerably.

The technique can be used to measure concentrations in the range 10“®-
10“®M and as such is eminently suitable for the determination of trace metal
impurities; of recent years it has found application in the analysis of semi-
conductor materials (Ref. 18) and in the investigation of pollution problems.

XVI, 23. SOME FUNDAMENTAL FEATURES. In view of the limitations
referred to above, and particularly the influence of electrode characteristics upon
the peaks in the voltammogram, some care must be exercised in setting up an
apparatus for stripping voltammetry. The optimum conditions require

(a) in the concentration step; a small mercury volume by comparison with the
volume of solution to be electrolysed; efficient stirring of the solution during
the electrolysis, otherwise the deposition procedure may be unduly
prolonged.

(b) in the stripping operation, as fast a voltage scan as possible consonant with
the avoidance of peak tailing.

Electrodes. The Hanging Mercury Drop Electrode is traditionally
associated with the technique of stripping voltammetry and its capabilities were
investigated by Kemula and Kublik (Ref. 19). In view of the importance of drop
size it is essential to be able to set up exactly reproducible drops, and this can be
done by attaching the capillary tube at the end of which the drops are to be
produced, to a small reservoir containing mercury and from which mercury can
be expelled in controllable quantities by means of a plunger operated by a
micrometer screw gauge. Problems which may be encountered with the HOME
include: (i) diffusion of the amalgam formed in the concentration process into the
capillary, with the result that it is not readily available during the stripping
process and the relationship between peak current and concentration breaks
down; (ii) possible penetration of aqueous solution into the capillary, which may
lead to the drop breaking away; this eventuality may to some extent be guarded
against by giving the bore of the capillary a coating of silicone before filling it with
mercury. Both of these problems can also be mitigated by using a Sessile Mercury
Drop Electrode (SMDE); in this variation, the end of the capillary tube is bent
upvyards through an angle of 180° so that the mercury drop sits on top of the
capillary instead of hanging from the end.

In an alternative technique, an HDME is set up on the end of a short platinum
wire sealed into a glass tube. The wire is first thoroughly cleaned and is then used
as anode for the electrolysis of a solution of pure perchloric acid: this treatment
exerts a polishing effect. The current is then reversed, and the electrode used as
cathode to ensure that no oxide or adsorbed oxygen films remain on the surface of
e electrode. Still using it as cathode, the electrode is now used for the electrolysis

665

XVI, 23 QUANTITATIVE INORGANIC ANALYSIS

of mcrcuryfll) nitruie soUnion, and it thus becomes plated with mercury. A
counted number of mercury drops from a conventional dropping mercury
electrode can then be attached to the platinum wire.

If the same procedure is carried out u.sing a longer platinum wire, and the
eleclrolysi.s conditions (current and time) with the mercury(II) nitrate solution
carefully controlled, then a Mercury Film Electrode (MFE) is produced. This is
cliiimed' to possessad vantages related to its rigidity and also to the greater surface
area/volume ratio as compared with a mercury drop. With both the electrodes
based on platinum wires, however, should there be any bare platinum areas
which have escaped plating in contact with the solution, then complications may
arise owing to the smaller hydrogen overpotential on platinum than on mercury,
and metals having high positive electrode potentials may fail to deposit when the
electrode is used for the concentration step.

Of recent years the use of mercury film electrodes based on .substrates other
than platinum has been explored, and increased .sensitivity is claimed for
electrodes based on wax-impregnated graphite, on carbon paste and on vitreous
carbon; a technique of simultaneous deposition of mercury and of the metals to
be determined has also been developed. For further details the review article{Rcf.
10) may be consulted.

Ce/N. The celt etnployed can be a suitable polarographic cell, or can be
specially constructed to fulfill the following requirements. Elficient reproducible
stirring of the solution is essential, and for this purpoic a magnetic stirrer is
usually suitable. Exclusion of o.xygen is important, and so the cell mast be
provided with a cover, and provision made for pa,ssing pure (o.xygen-frce)
nitrogen llirough the solution before commencing the experiment, and over the
surface of the liquid during the determination. The cover of the ceil must provide
a firm seating for the UMI.)E (or other type of electrode used), and must also have
openings for liie reference electrode (usually a SCE) and for a platinum counter
electrode if it is requited to operate under conditions of controlled potential (see
Section X\T, 8). If the solution under investigation is to be analysed for mercury,
then the reference electrode .should be isolated from the solution by use of a .salt
bridge.

liciii^ciu.s. In view of the sensitivity of tlic method, the reagents employed
for preparing the ground solutions must be very pure, and the water used should
be redistilled in an all-gl.is.s apparatus: the trace.s of organic material sometimes
encountered m demineralised water (Section 111, 21 ) ma)ce such water unsuitable
for this technique. The common supporting electrolytes include pota.ssium
chloride, sodium acetate acetic acid bulTcr solutions, amraonia- ammonium
chloride buffer solutions, hydrochloric acid and potassium nitrate.

fhe normal A.R, grade chemicals often contain trace impurities which arc
quite unimportant for most analytical purposes, but in terms of stripping
voltammetry may represent scriou.s contamination: this is especially true if heavy
melais are involved. It is therefore nccc.s,sary to employ reagents of very high
purity (e.g., the B D.ll. ‘Aristar’ reagents or .similar grade), or alternatively to
subject the purest material available to an electrolytic purilitxition process. The
stirred solution is electrolysed with a small current ( 10 m.A) for twenty-four hours,
using a pool of mercury at the bottom of a beaker as cathode and a platinum
anode; pure nitrogen is passed through the solution before commencing the
electrolysis so iis to remove dissolved o.xygen, and during the purification process,
a current of pure nitrogen is maintained over the surface of the solution. It will

666

VOLTAMMETRY XVI, 24

usually be necessary to use some form of potentiostatic control during the
electrolysis process. When electrolytic purification of reagents needs to be
undertaken on a routine basis it may be considered advisable to make use of
commercially available apparatus such as, for example, The Princeton Applied
Research Model 9500 Electrolyte Purification Apparatus.

In view of the foregoing remarks, it is clear that all glassware used in the
preliminary treatment of samples to be subjected to stripping voltammetry, as
well as the apparatus to be used in the actual determination, must be scrupulously
cleaned. It is usually recommended that glassware be soaked for some hours in
pure nitric acid (6M), or in a 10 per cent solution of pure 70 per cent perchloric
acid.

XVI, 24. INSTRUMENTATION. Many of the more sophisticated
polarographs referred to in Section XVI, 20 are suitable for carrying out stripping
voltammetry and some manufacturers supply equipment specifically designed for
this technique: amongst these may be mentioned the ‘ElectRoCell-ASV’
apparatus supplied by McKee-Pedersen Instruments which incorporates a
rotated cell and by this means achieves the controlled stirring of the solution
which is required, and the Environmental Sciences Associates Inc. Anodic
Stripping Voltameter. The latter can be purchased in single cell (Model SA2011)
and multi-cell (Model 2014) versions: the standard (20 mm) cell holds a working
volume of about 5 cm^ and stirring is accomplished by a stream of pure nitrogen.

Fig. XVI, 21

w St the standard electrode assembly consists of a mercury-coated, wax-
impregnated graphite rod, a Ag/AgCl reference electrode and a platinum counter
ec rode. The multiple cell model (2014) together with a four-cell holder is shown

in rig. XVI, 21.

667

XVI, 25 QUANTITATIVE INORGANIC ANALYSIS

Chronopotentionietry

XVI, 25. B.VSIC PRINCIPLES. If an un.stirrcd ground solulion coniaininga
small amount of a dcpolariser is cIcclroiy.scd at constant current using a cell
provided with a working electrode, a counter electrode, and a reference electrode
(cf. Fig. .XllI, 2; Section XIII, 2), and the potential of the working electrode is
plotted against lime, then the resultant curve is similar to a conventional
polaroeram (Fig. XVI, 3; Section XVI, 2). The curve is referred to as a
clironopotcnfiogram and i(.s shape can be e.xplained as follows.

As shown in equation ( .^) (Section XVI, 3), the potential of a cathode at which
a reversible reduetion reaction is involved

Ox -f lu; Red
is given by

E ~ E''’ + RT;'«F In [Ox];tRcdJ.

As soon as a small amount of reductant has been produced, the ratio (0.xl/[Red]
only changes relatively slowly, and hence the potential of the electrode changes
only gradually with respect to time. As electrolysis procced.s, however, the
concentration of the oxidant adjacent to the electrode decreases, and although it
is to a certain extent replenished by dilTu.sion. if the magnitude of the electrolysis
current is high enough, a situation is reached in which the concentration of the
oxidant in the electrode layer is virtually reduced to zero. The conditions for a
fixed electrode potential no longer apply, and the potential changes rapidly to a
value at which ;i new electrode reaction is po.ssible. The time from the
commencement of electrolysis to the rapid change in potential is termed the
transition time t, and it was shown by Sand (Kef. 20) that this is related to
the concentration of the electro-active sj>ecies by ilie expression

^ if

where ii ~ number of electrons involved in the reduction reaction,

A == surface area of electrode.

D = dilTusion coclficienl of the clcciro-activc sjK'cics involved,

Co ~ initial concentration of the dcpolari-scr,

/ = the consianieleciroly.sis current.

The variation of electrode potential with time can be expressed by the equation

C- £'■•’ + — In

and when l ~ r. 4. the potential F. ^ (the quarter (raasition time potential) ^ —

£,2 (Ihc polarographic half wave potential).

It follows from tiie.so two equations that, like polarography, chronopotentio-
mciry has potentialities for use in both quantitative and qualitative analysis, but
from the quantitative viewpoint it is a less .sensitive technique than others already
described in this chapter, and the lower limit of concentration which can be
mea.surcd with reasonable accuracy is about 10 "'*3/. Complications also arise

668

VOLTAMMETRY XVI, 26

when the solution under investigation contains more than one reducible species.
For the first one to be reduced, the transition time (tj) will be given as above by

but the transition time (T 2 ) for reduction of the second species is given by
rr^l'^n TAD ^1'^ C

and if a multivalent ion.is reduced in two stages involving respectively and nj
electrons, then

(ri+T2)/Ti =(ni + «2)V«i^-

This interdependence of transitions times is clearly disadvantageous as
compared with the normal polarographic situation where the height of any wave
is proportional to the concentration of the species giving rise to it and is quite
independent of the height of the preceding waves.

The fact that there is no periodicity in the measurements such as that
associated with the use of a DME is sometimes regarded as a point in favour of
chronopotentiometry, but on the whole, although the technique has been applied
to quantitative determinations (Ref. 21), its main uses lie in the investigation of
the kinetics of electrode reactions, in confirming (or disproving) their
reversibility, in ascertaining the number of electrons involved in the reaction, and
in establishing the formulae of complexes (Ref. 22).

XVI, 26. EXPERIMENTAL PROCEDURE. The working electrode is
commonly a pool of mercury, but it can be a hanging mercury drop, a wax-
impregnated graphite rod, or a platinum disc sealed into a glass tube and situated
a few millimetres from the end of the tube: this arrangement helps to prevent
convection effects. The reference electrode (most commonly a SCE or a Ag/AgCl
electrode) is clamped firmly in position and is provided with a drawn-out tip
which is situated close to the working electrode. The auxiliary (counter) electrode
(a small platinum plate or a wire helix) is usually separated from the solution to be
analysed by placing it inside a glass tube fitted at the end with a sintered glass
diaphragm; this tube contains the pure ground solution.

A polarographic cell similar to that depicted in Fig. XVI, 13 (Section XVI, 9)
roay be used but with the counter electrode placed in the right-hand

compartment, the DME replaced by the reference electrode, and an electrical
connection provided to the mercury pool at the bottom of the left-hand
compartment; this is the working electrode. Provision must also be made (as in
he cell depicted) for passing pure nitrogen through the solution to remove
issolved oxygen before commencing the electrolysis.

The constant current for the electrolysis can be obtained from a battery in
senes with a variable high resistance, but is best produced by a commercially
available amperostat such as the Solartron, Beckman ‘Electroscan’, McKee-
e ersen, Princeton Applied Research, etc.

se of a chart recorder enables the voltage-time plot to be produced directly
n It IS usually considered that the current should be adjusted so that the
nsi ion times lie between 10 and 100 seconds; the procedure described in

669

XVI, 27/28 QUANTITATIVE INORGANIC ANALYSIS

Section XVI, 6 can be employed lor precise evaluation of the transition times
from the graph.

XVI, 27. References

1. D. R. Crow aiul J. V. WosOvood (1968). Polciroj^utphy, London; Methuen and Co.
Ltd.

2. J. He>Tov.sky (1922). Chankkc Liny.. 16, 256,

.1. J. Hcyrovslty and M. Shikala (1925), Rcc. iriir. chim., 44, 496.

4. D. R. Crow 0969). Roturoyraphy ufMcinl Complexes. London; Academic Press.

5. J. B. Plato (1972). ‘The Rcnais;ance in Polarographic and Voltammeiric .-Vnaljsis'
(review article). Amtl. Client., 44, 75A.

6. B. Breycr, F. Guiinan and S. I lacobian (1950). .iustruUan J. Set. Res.. A3, 558.

7. D. llkovie (1934). CM Czech. Chem. Comm.. 6. 498.

8. G. C, Barker and 1. L. Jenkinv (1952). . t/;<//>.ir, 77. 685.

9. G. C. Barker and A. W. Gardner (1960). oiuil. Chem., 173, 79.

10. B. Fleet and R. D, Jee (1973), SpecitiliM RerioiUcol Reporis, ElectrochemLury. Vol. 3.
London ; The Chemical Socieiv

11. J. HevrovskvandJ. For!et(I943) X. Phs.s. Chem., 193.77.

12(a). J. F. B. Randles (1947). .tm;/wr, 72. 301.

12(b). J. E H. Randles (1948) /‘rum. Far .S‘<(e.. 44. .327; 334.

13(a). 11, M. Davisand J. E Seaborn (1953). Rlecinmie lioywecrhiit.lb, 314.

13(b). H. .M. Davis and J. E Seaborn {I960), .ithames in Rohtroyriiphy. \'ci|, 1. O.vford;
Pergamon Press,

14(a). R. Kalvoila (1965). rechnnjues of (halhiyr.iphic I'olaroyniphy, Amslerdani;
Elsevier.

14(b) R. Kalvuda. W. Anstine and Nt Ueyrovskv ( 1970). .Icii/, Chtm. .tetu., 50,93.
15(a). G. C, Barker. R L. Faircloth and A W Gardner) 1958). iVornre, 181, 247.

15(b), B. S, Brnk and B, .\1. .Siernber;e (1970) ZusoJ. IM>., 36, 365,

)6(a). P. E, Toren(19(.S) . inu/. C7ia;i..40, I !52

16(b). G. C. Wliilnackand R. G. Brophv {!9o9) ,■{/;, i/ Chim. .((■ru.,48, 123.

16(e). C. E Ploek .ind J. Vas4iuer(i971). -Ok.'/- C/ir«j .-Icr.r.. 55. 278.

17. Information Biillelin (1973), ApjKndiees on rem.uive Nomenelature. S>mbob,
Unil.s and Standards, No, 30. Clasn/ieiilion iitui Xomenchiiure of Eteciroanalyiicol
I'eehniipies. O.vford. lUPAC.

18. P F K.meandG, B L.irr.ibce ( 1970) CkoraetetisitiionofSemkonJuctor Materials.
New 'I'ork, .McGraw-Ilill Book C'o

19. W Kemulaand Z. Kub!ik(1958). .In.;/. Chtm. .ielii , 18. 101,

20 11 J S. Sand (1901) Phil. .May.. l.A>.

21. D. G. Davis in A. J, Bard (1966). Eleetroiiiuilytkal Chemistry. Vo). 1, New York; M.
Dekkcr Inc.

22. J. B, Uc.idndge (1969). Eleciruehemical Teehniques for Inoryanic Chetnisls. London;
Academic Press.

23. R. S Nicholson (1972). 'Polarogr.iphic Thcorv, Instrumentation and Methodology’
(review article). , Inal, ( hem., 44. 47SR.

XVI, 28. Selected bibliography

I. D. E. Burge (1970). ‘Pulse Polarography*, J. Chem. Ed.. 47. ASl.

2- D E. Smith, o.c. Pohiroyraphy. in A 3. Bard (Ref. 21).

3. A. J. Bard (1967). Eleetroinutlytical Cliernistrv. V'ol. 2. IStrippiny Voltammclry:
Oscillographic Polaroyrapln) New York; M. Dekkcr Inc.

4. J. Heyrovsky and ). )Cuta (1966). Prineiptes of Polurograpliw New York; Academic
Press.

670

VOLTAMMETRY XVI, 28

5. G. Chariot, J. Badoz-Lambling and B. Tremillon (1962). Electrochemical Reactions.
Amsterdam ; Elsevier.

6. J. Heyrovsky and P. Zuman (1968). Practical Polarography. New York; Academic
Press.

7. O. P. Bhargava, W. D. Lord and W. G. Hines (Sept./Oct. 1975). Anodic stripping
voltammetric determination of lead and zinc in iron and steelmaking materials.
International Laboratory. Fairfield, Conn. ; International Scientific Communications
Inc.

CHAPTER XVII AWIPEROIVIETRY

XVII, 1, AMPKUOMEIRIC TITRATIONS. It has been shown in the
previous Chapter (Section XVI, 3; Fig. XVI, 4) that the limiting current is
iiuie{)eniient of the applied \ oltage impressed upon a dropping mercury electrode
(or other indicator micro-electrode). The only factor alVecling the limiting
current, if the migration current i.s almost eliminated by the addition of suflkieni
supporting electrolyte, is the rate of dilfusion of electro-active material from the
bulk of the solution to the electrode surface. Hence the dilTusion current
{ - limiting current - residual current ) is proportional to the concentration of the
elccira-aciivc malcnal in the solution. If some of the electro-active material is
removed by interaction with a reagent, the dillu.sion current will decrease. This is
the fundamental princi ple of ampciomclric titratio ns. The observed diffusion
current at a suitable applied voltage is measured as a function of tiie volume of
the titrating .solution: the en d p oint is the point of inte r.section of two lines giving
the change of currentbeibfe and^tcr the cquiVMicnccpiHhf. - ■ •

U may be noted that when during a titration the potential is measured between
an indicator electrode and a referenee clectroile, the titration is termed a
potentionielric one: here it is important to measure the potential relatively
accurately near the end point, the latter being characterised byama.\imumorthe
dilTereniial .Aif/Av, ihe rale of change of potential. s.iy, per 0.1 cm\ in
conductometric titrations the electrical conductivity of the solution is meu.sured
during the titration, and the end point is found graphically as the point of
intersection of two .straight lines giving the change of conductivity l>cfore and
after the eciuivalenec point (compare high-frequency titration methods). In
amperomelric titrations (derived from ampere, the unit of current) the current
which passes ilirough ilie liiraliou cell between an indicator electrode (e.g.. the
dropping mercury electrode) and the appropriate depolarised reference electrode
(e.g., the saturated calomel electrode) at a suitable applied c.m.f. is measured as a
function of the volume of the titrating solution. .Such titrations have also been
termed polarographic and polarornciric; the term umpgrrmietric titration is now
recommended. ^

Some advantages of amperometric titrations may be mentioned:

1. The titration can usually be carried out rapidly, since the end-point is found
graphically; a few current measurements at con,stant applied voltage before
and after the end point siillicc.

2. fitraiions can be carricv! out in cases in which the solubility relations arc such
that potentionielric or visual-indicator methods arc unsatisfactory; for

672

AMPEROMETRY XVII, 1

example, when the reaction product is markedly soluble (precipitation
titration) or appreciably hydrolysed (acid-base titration). This is because the
readings hear the equivalence point have no special significance in
amperometric titrations. Readings are recorded in regions where there is
excess of titrant, or of reagent, at which points the solubility or hydrolysis is
suppressed by the mass-action effect; the point of intersection of these lines
gives the equivalence point.

3. A number of amperometric titrations can be carried out at dilutions {ca.

at which visual or potentiometric titrations no longer yield accurate
results. (It must be noted, however, that high frequency titrimetry may also be
applied to dilute solutions [see Chapter XV]; the potentiometric method is
superior for more concentrated solutions.)

4. ‘Foreign’ salts may frequently be present without interference and are, indeed,
usually added as the supporting electrolyte in order to eliminate the migration
current.

5. The results of the titration are independent of the characteristics of the
capillary.

6. The temperature need not be known provided it is kept constant during the
titration.

7. Although a polarograph is convenient as a means of applying the voltage to
the cell, its use is not essential in amperometric titrations. The constant
applied voltage may be obtained with a simple potentiometric device (see Fig.
XVI, 9; Section XVI, 7).

If the current-voltage curve of the reagent and of the substance being titrated
are not known, the polarograms must first be determined in the supporting
electrolyte in which the titration is to be carried out. The voltage applied at the
beginning of the titration must be such that the total diffusibn current of the
substance to be titrated, or of the reagent, or of both, is obtained. In Fig. XVII, 1,
are collected the most common types of curves encountered in amperometric
titrations together with the corresponding hypothetical polarograms of each
individual substance : S refers to the solute to be titrated and R to the titrating
reagent. The slight ‘rounding off in the vicinity of the equivalence point is due to
the solubility of the precipitate; this curvature does not usually interfere, since
the end-point is located by extending the linear branches to the point of
intersection. For each amperometric titration the applied voltage is adjusted to a
value between X and Y shown in Fig. XVII, 1, A'-D \ In A onl y the material
being titrated gives a diffusion current (see A'), i.e., the electro-active material is
removed from the solution by precipitation with an inactive substance (for
example, lead ions titrated with oxalate or sulphate ions). In B the solute gives no
diffusion current but the reagent does (see B'), i.e., an electro-active precipitating
reagent is added to an inactive substance (for example, sulphate ions titrated with
barium or lead ions). In C both the solute and the titrating reagent give diffusion
currents (see C') and a sharp V-shaped curve is obtained (for example, lead ion
dichromate ion, nickel ion with dimethylglyoxime, and copper ion
with betizoin a-oxime). Finally, in D the solute gives an anodic diffusion current
(hat IS, is oxidised at the dropping mercury cathode) at the same potential as the
1 rating reagent gives a cathodic diffusion current (see D'); here the current
c anges from anodic to cathodic or vice versa and the end-point of the titration is
in mated by a zero current. Examples of D include the titration of iodide ion with
ercury(Il) (as nitrate), of chloride ion with silver ion, and of titanium(III) in an

673

AMPEROMETRY XVII, 2

acidified tartrate medium with iron(III). Became the diffusion coefficient of the
reagent is usually slightly different from the substance being titrated, the slope of
the line before the end-point differs slightly from that after the end-point
(compare D); in practice, it is easy to add the reagent until the current acquires a
zero value or, more accurately, the value of the residual current for the
supporting electrolyte. ' ;

To take into account the change in volume of the solution during the titration,
the observed currents should be multiplied by the factor (V+v)jV, where V is the
initial volume of the solution and v is the volume of the titrating reagent added.
Alternatively, this correction- may be avoided (or considerably reduced) by
adding the reagent from a semi-micro burette in a concentration ten to twenty
times that of the solute. The use of concentrated reagents has the additional
advantage that comparatively little dissolved oxygen is introduced into the
system, thus rendering unnecessary prolonged bubbling with inert gas after each
addition of the reagent. The migration current is eliminated by adding sufficient
supporting electrolyte; if necessary, a suitable maximum suppressor is also
introduced.

^Vn, 2. TECHNIQUE OF AMPEROMETRIC TITRATIONS WITH THE

^ DROPPING MERCURY ELECTRODE. An excellent and inexpensive
titration cell consists of a commercial resistance glass (e.g., Pyrex), 100-cm^,
three-necked, flat or round-bottomed flask to which a fourth neck is sealed. The
complete assembly is depicted schematically in Fig. XVII, 2, A. The burette
(preferably of the semi-micro type and graduated in 0.01 cm^), dropping
electrode, a two-way gas-inlet tube (thus permitting nitrogen to be passed either
through the solution-or over its surface), and an agar-potassium salt bridge (not
shown in the figure) are fitted into the four necks by means of rubber stoppers.

The agar-salt bridge is connected through an intermediate vessel (a weighing
bottle may be used) containing saturated potassium chloride solution to a large
saturated calomel electrode. Th e agar-salt bridge is ma de fro m a gel which is ^
per cent j n agar and contains sufficient pota ssium chionflfi fn caturQ^A
soluiion__a^e room temperaju^when Thlmi'dh' ions interfere with the
itrations, the connectionjsjuade with an agar-potassium nitrate bridge.

675

XVII, 3 QUANTITATIVE INORGANIC ANALYSIS

Another cell, due to Ling ane and Laitine n, is shown in Fig. XVH, 2, B; the
special feature of this FFcellTs the .sintered-glass disc (porosity 3) and the 3 per
cent agar-s;ilt plug whieh separates the saturated calomel electrode from the
solution being titrated. A minor dJ.sadv antape would app ear t o be the po ssibilitv
of breaking the fragile capillary or tluTljuretlc lip upon reino yal from the rubber
stqppert.-tf--dcsiriS.rrnFe ngITFhaifd"cminiaflihenf c;rn"'Fe Tilled tvuF'saiurated
potassium chloride .solution and connection with the e.\ternal reference electrode
made with another salt bridge in the usual way.

. ■- Thermostatic SQJ3itdis,xicixss<muaipmvided;btK-e!H3-niatntituwtkauUjairli;_
constant temper.iture during the titration. It is advantageous to sto re the rcai'cni
bencathlTn-atraosphere, of inert easLtiutt,nfec aution is not ah sc tlutely necessary if
the reagent solution has ten to twenty times the concentration of the solution
being titrated and is added from a semi-micro burette. If the solute is clectro-
' reducible. suHicicni electrol>tc should be added to eliminate the migration
current; if the reagent Ls electro-reducible and the solute is not, the addition of a
supporting electrolyte is usually not required, since .sullieienl electrolyte is
formejJ during the titration to eliminate the migration eurrent beyond the end
point, h’niay be necessary to add a suitable maximum suppressor, such as
gelatin. If the polarographic ciiaracteri-stics of the solute and the reagent are non
known, the current- voltage curse of each must be determined in the medium ini
'w hieh the titration is being carried out. The applied voltage is then adjusted at the
beginning of the titration to such a value that the dilfu-sion current of the
unknown sulutc, or of the reagent, or of both, is obtained ; frequently the voltage
range i.s comparatively large and, in consequence, great accuracy is not required
in adjusting the applied voltage. ~

The geiier.il procedure i.s as follows. A known volume of thesolution under lest
is placed in the titration cell, whieh is then ;issenibled as in Fig. XVII, 2, A: the
electrical conncx'tions are completed (dropping mercury electrode as cathode;
saturated calomel half-cell, or mercury pool at bottom ufllask, as anode), and
dissolved o.xygen is removed by p.i.ssing a slow stream of pure nitrogen forabout
15 minutes. The applied voltage is then adjusted to the desired value, and the
initial dilfusion current is noted. A known v olumc of the reagent is run in from a
.semi-micro burette, nitrogen is bubbled through the solution for about 2 minutes
to eliminate traces of oxygen from the added liquid and to ensure complete
mixing. The Iknv of gas liiituutli the solution is then stopped, but is allowed to
pass over the surface of the solution (thus maintaining an inert, o.xygcn-free
atmosphere). The current and burette readings are both noted. This procedure is
repealed until suiiiciein readings have been obtained to permit the end point to
be determined as the inlcr.section of the two linear parts of the graph.

XVU. 3, DE I’EUMlN.VnON OF LEAD WIT 1 1 ST ANDARD POT'.VSSIUM
DICHROM.ATE SOLUTION. Both lead ion and dichromaie ion yield a
dilfusion current at unapplied potential to a dropping mcrcuryelectrodeof — I.O
volt ag;iinst the satuniled calomel electrode (SCE). Amperometric titration gives
a V-shaped curve (Fig. XV 1 1, I, C). For convenience in its use by large classes ol
students, the exercise has been adapted to the determination of lead in A.R. lead
nitrate; the application to the deierinination of lead in dilute aqueous solutions
( 1 0 ' ''~1 0 ■ ■*3/) is self-evident.

Reagents required. Dissolve an accurately weighed amount of A.R. lead
nitrate in 250cm^ water in a graduated flask to give an approximately 0.013/

676

AMPEROMETRY XVII, 4

solution. For the titration, dilute 10 cm^ of this solution (use a pipette) to 100 cm
in a graduated flask, thus yielding a ca. O.OOIM solution of known strength.

Prepare a ca. 0.05M solution of potassium dichromate using the appropriate
quantity, accurately weighed; of the d^ A.R. solid. Dilute this solution to ca.

0.005M. . ' • . . ■

Prepare also a ca. O.OlAf solution of potassium nitrate from the A.R. solid for
use as the supporting electrolyte.

Procedure. Use any commercial or manual polarograph: see Chapter

XVI. Set up the dropping mercury electrode assembly and allow the mercury to

drop into distilled water for at least 5 minutes. Meanwhile, place 25.0 cm^ of the
ca. O.OOlM-lead nitrate solution in the titration cell, add 25 cm^ O.OIM-
potassium nitrate solution, complete- the cell assembly, and bubble nitrogen
slowly through the solution for 15 minutes. Make the necessary electrical
connections. Apply a potential of — 1 .0 volt vs. SCE : at this potential both the
lead and the dichromate ions yield diffusion currents. Turn the three-way tap so
that the nitrogen now passes over the surface of the solution. Adjust the
galvanometer sensitivity so that the spot is on the scale and take the reading. Do
not alter the applied voltage during the determination. Add the ca. 0.005M-
dichromate solution in 0.5-cm^ portions until within 1 cm^ of the end point, and
henceforth in 0.1 cm^ portions until about 1 cm^ beyond the end point, and
continue with additions of 0.5 cm^. After each addition pass nitrogen through the
solution for 1 minute to ensure thorough mixing and also deoxygenation, turn
the tap so that the nitrogen passes over the surface of the solution, and note the
deflection of the galvanometer spot, i.e., measure the current. It will be observed
that a large initial current will decrease as the titration proceeds to a small value
at the equivalence point, and then increase again beyond the equivalence point.'
Correct the readings of the galvanometer deflection for the change in volume of
the solution due to the added reagent using the formula (K+y)/F,

where V is the initial volume of the solution and v is the volume of the titrating
reagent. Plot the values of the corrected current (galvanometer deflections) as
ordinates against the volume of reagent added as abscissae: draw two straight
lines through the branches of the ‘curve’. The point of intersection is the
equivalence point. Calculate the percentage of lead in the sample of lead nitrate.

1 cm^ 0.01M-K2Cr207 = 0.002072 g Pb

Repeat the titration using 0.05M-dichromate solution added from a 5- or 10-
cm^ semi-micro burette.

XVII, 4. DETERMINATION OF SULPHATE WITH STANDARD LEAD
NITRATE SOLUTION. Solutions as dilute as O.OOIM with respect to
sulphate may be titrated with O.OlM-lead nitrate solution in a medium
containing 30 per cent ethanol with reasonable accuracy. For solutions 0.0 1 M or
higher in sulphate the best results are obtained in a medium containing about 20
per cent ethanol. The object of the alcohol is to reduce the solubility of the lead
sulpha te and thus minimise the magnitude of the rounded portion of the titration
curve in the vicinity of the equivalence point. The titration is performed in the
absence of oxygen at a potential of - 1.2 volts (vj. SCE) at which potential lead
ions yield a diffusion current. A ‘reversed L’ graph (compare Fig. XVII I B) is
obtained : the intersection of the two branches gives the end-point. A supporting
electrolyte need not be added, since the current does not increase appreciably

677

XVll, 5 QUANTITATIVE INORGANIC ANALYSIS

until an excess of lead is present in the solution, and the amount of salt formed
during the titration suflices to completely suppress the migration current of lead
ions.

Reagents required. Prepare an appro.ximately 0.01 d/ solution of
potassium sulphate in a 100 -cm^ graduated flask using -an accurately weighed
quantity of the dry A.R. solid. Similarly prepare an approximately O.lAMead
nitrate solution in'a 100 -cm^ graduated fla-sk from a known weight of the dry
A.R. solid.

Frucciltin-. U.se the apparatus and technique described in the previous
Section. Introduce 25,Ocnv‘ of the pota.ssium sulpliale solution into the cell, add
2 to 3 drops of thymol blue followed by a few drops of concentrated nitric acid
until the colour is just red (pH 1.2); finally, add 25 enu* of 95 [x-r cent ethanol.
Connect the saturated calomel electrode through an agar--poiassium nitrate
bridge to the cell. Pill ll>e semi-micro burette with the .standard lead nitrate
solution. Pasi nitrogen through the solution in the cell for 15 minutes and then
over tlie surface of the solution. Meanwhile adjust the applied voltage to - 1.2
volt. Set the seibitiviiy control at the appropriate value and also the
galvanometer spot at zero. Introduce the lead nitrate .solution from the burette in
0.5-cm^ portions until within Icm^ of the equivalence point, then in O.l-cm^
quantities for the following 2 cm^, and subsequently in 0.5-cm^ portions. Pass the
gas stream through the solution for about I minute after each addition (more
dilute solutions will require up (o 3 minutes to assist the precipitation of the lead
sulphate) and then over the surface before reading the galvanometer deflection
(current). Correct the current readings for tire cliange in volume of the solution
due to the added reagent a.s in the previous c.xperiment. Read oflThe equivalence
point from the amperoinctric titraiimi curve drawn from your results.

Calculate the (vercctitage of SO 4 in the sample of A.R. potassium sulphate.

1 cm^ O.I.l/.PblNOj), ().(J(J9()(l6g.S'Oi

.WII, 5. DETKR.MIN.vnON OI‘ NICKEL WITH UIMETHYLGLY-
OXIME. The nickel solution (concentration less than 0,005.\/) is intro-
duced into an aqucou.s ammonia-ammonium chloride supporting medium
and. after deoxygenalion. titration is carried out at an applied voltage
(-I.S5 volts I’.v. SCE) at which both nickel and dimethylglyoximc arc
reducible. A V-shaped titraliou graph (Fig. XVI !. 1 , C) is obtained.

Reagents required, (i) Prepare a 0.02.f/-dimethylglyo,xime solution by
di.ssolving 0.2322 g of A.R. dimethylglyoximc in 95 percent ethanol (rectified
spirit) and make up to 100 cm ’ in u graduated flask with the same solvent.

(ii) Prepare an approximately 0.013/ solution of ammonium nickel sulpliate
by weighing out about 0.395 g of the salt (preferably of A.R, quality) and
dissolving it in 100 cm-* of vvmer m a graduated flask. Standardise the solution by
an ED I A titration (Section X, 58). Dilute 25.0 cm^ of thi.s solution to 250cni^ in
a graduated flask, thus giving a ai. 0.0013/ solution.

(iii) Prepare the base solution by dissolving d.Ocnf^ ofconcentrated ammonia
solution (sp.gr. 0.88) and 5.35 g of A.R. ammonium chloride in water and
diluting to 1 dni^ in a graduated flask. The resulting .solution is 0.53/ in aqueous
ammonia and 0.1 3/ in ammonium chloridc.

Piwcilure. Use tlie four-necked lOO-cnT* titration flask depicted in Fig.
XV! 1, 2, A including an agar-potassium nitrate bridge and a microburcltc. Place

678

AMPEROMETRY XVH, 6

25.0 cm^ of the ca. 0.001 A/-nickel solution, 25 cm^ of the base solution, and 1 cm^
of 0.2 per cent gelatin solution in the clean, dry titration vessel; the base solution
will now be ca. 0.25M in aqueous ammonia and ca. 0.05M in ammonium
chloride. Pass oxygen-free nitrogen through the solution for 15 minutes. Raise
the dropping mercury electrode reservoir and allow the mercury to drop into
distilled water for 5 minutes. Meanwhile connect a. saturated calomel electrode
through an intermediate saturated potassium chloride solution by means of an
agar-salt bridge to the titration vessel. Fill the semi-micro burette with the
0.02M-dimethylglyoxime solution and insert the tip inside the titration flask.’ .

Set the applied potential at — 1.85 volts versus the saturated calomel electrode,
commence the flow of mercury from the dropping electrode and note : the
maximum deflection of the galvanometer spot. Add the dimethylglyoxime
solution from the semi-micro burette in suitable increments (e.g., of 0.2 cm^)
until within 1 cm^ of the end-point; then reduce the additions to 0.05-0.1 cm^
and continue well beyond the equivalence point. After each addition pass
nitrogen through the solution for 1 minute to deoxygenate and to mix the
solution, and then observe the galvanometer deflection (current). It will , be
observed that the galvanometer deflection (current) decreases linearly to the end-
point and then increases more rapidly. Plot current (ordinates) against volume of
dimethylglyoxime solution (abscissae), making the appropriate correction for the
volume of reagent added at each reading. The equivalence point is the point of
intersection of the two linear branches of the graph. Calculate the percentage of
nickel in the sample of ammonium nickel sulphate.

1 cm^ 0.02M-dimethyIglyoxime = 0.0005869 g Ni

Note. In the above determination, the dimethylglyoxime is assumed to be
pure. It is better to check the purity of the dimethylglyoxime with a standard
nickel solution and to use the resulting factor for the dimethylglyoxime solution
in subsequent calculations. Then determine the nickel content of a solution
containing 0.05-1 mg of nickel. Other elements which form complexes with
dimethylglyoxime, especially cobalt, copper, and bismuth, must be absent.

Many other metals can be similarly determined by amperometric titration with
suitable organic reagents; a full selection is given in Ref. 10.

XVn,6. DETERMINATION OF FLUORIDE WITH STANDARD
THORIUM NITRATE SOLUTION. Neutral solutions of fluoride may be
titrated with 0.01 M-thorium nitrate in a medium of 0. IM-potassium chloride at
an apphed potential of — 1.7 volts vj. SCE. Thorium ions are not reducible at the
dropping mercury cathode, but they seem to have the property of carrying to the
mercury cathode nitrate and nitrite ions which are reduced, producing a step with
a half-\yave potential of — 1.3 volts. The height of this step is roughly
proportional to the concentration of thorium ions in solution, consequently a
reversed L-type of titration graph is produced in the titration of fluoride.

Reagents required, (i) Prepare a ca. O.OlM-thorium nitrate solution by
dissolving about 5.8 g A.R. thorium nitrate in 1 litre distilled water. The solution
may be standardised against a standard fluoride solution by amperometric
titration.

(ii) Prepare a standard O.OIM fluoride solution by dissolving 0.1050 g
accurately weighed, dry A.R. sodium fluoride in 250 cm^ water in a graduated
flask. Transfer 25.0 cm^ of this solution to a 250 cm^ graduated flask containing

679

XVn,7/8 QUANTITATIVE INORGANIC ANALYSIS

1 .86 g A.R. poUKS-sium cliloridt;. Sliaicc until dissolved and dilute to the mark with
distilled water. The resulting solution is 0.001.4/ in lluorideand is also 0.1.1/ with
respect to potassium chloride.

Procedure. Check the pH of the fluoride solution with a pH meterr it
should be in the range 7"8. Place 25.0cm-’ of the neutral standard fluoride
solution in the titration flask (I-ig. XVll, 2, A), set the applied voltage at - 1.7
volts vs. SCIi, and titrate with the thorium nitrate solution in the usual manner.
Plot the titration curve, and evaluate the e.xact concentration of the thorium
nitrate solution.

Repeal the titration with an ‘unknown' neutral fluorklcionsolulion. say, ofeu.
0.0005 jU concentration; the ba.se electrolyte should be 0.1.1/ in potassium
chloride.

I cm-’ 0.01 .IZ-TlKNOj).! h ().0007600g F~

.XVH, 7. DE TERMIN.ATIO.N OF ZI.NC WITH EDI'A. Zinc ions may be
titrated with standard HDTA solution in a strongly alkaline medium (produced
with cyclohexylamine) at an applied potential of- 1.4 soils rs. SCh‘. Under these
conditions the diifusion current due to /inc ions decreases during the titration
and an L-slvapcd titration graph results.

Reagents required, (i) Prepare a standard 0.02.1/ /.inc-ion solution by
dissolving about 1.31 g. accurately weighed. .A.R. rinc in dilute hydrochloric acid
and diluting to 1 diiri with distilled water in a graduaieil flask. Dilute 25.0cni^ of
this solution to lOUcm’ in a graduatetl flask, thiis giving a at. 0.005.1/ zinc-ion
solution.

(ii) Preparea standard 0.01.l/-I;DTA solution (Section 50).

Procedure. Place 5.0()cm’ of the /inc-ion .solution in the titr.ition flask,
add 1.0 enr’ pure cyclohe.vylamine and 19.0 cm-* distilled water. Set the applied
potential at -1.4 volts v.w SCE. Deaerate the solution and titrate with the
standard EDTA solution in the asiml manner. Plot the titration graph, evaluate
the eonccniration of the /me in the solution, and compare it with the known
value.

Repeat the titration using an 'unknown' solution of /,iac ions, say, ofO.OdOj.l/
concentration.

1 cnF 0.OI.I/-EDTA ().(HJ0{»5.38g Zn

.XVIl, 8. TiTRATlC)N OF AN lODiUE SOLUTION WITH .MERCURY(II)
■NT riLATIv .SOI.UTION. This c.vperimcut iliustratc.s the titration of a
substance yielding an anodic step (iodide ion) with a solution of an oxidant
(mercury(Il) nitrate) giving a cathodic diifusion current at the same applied
voltage. '1 lie magnitude of the anodic diifusion current decreases up to the end-
point : upon adding an exce-ss of tiiraiu, the dilTusion current increases, but in the
opposite direction. The type of graph obtained is similar to that in Fig. XVil, 1,
D. The end-point of the titration is given by the iiiterseeiion of the two linear
portions ol the graph with the volume (of titrant) axis: ilie dilTusion current is
then approximately zero. Tlie two linear parts do not usually iiave the.saineslQ[X‘,
because tlic litraiu and the subsianee being titrated have diflerent dilTusion
currents for cquivaleiu concentratioms.

Reagciit.s rei|iurcd. (i) Prepare a ca. 0.004,l/-potassiuiu iodide solution by
dissolving 0.68 g A.R. potassium iodide, accurately weighed, in 1 dm^ water.

680

AMPEROMETRY XVII, 9

(ii) Prepare a O.OlM-mercury(II) nitrate solution, by dissolving 1.713,g pure
mercury(II) nitrate monohydrate in 500 cm^ of 0.05M-nitric acid.

(iii) O.lif-nitricacid.

Procedure. Equip a lOO-cm^ four-necked flask (compare Fig. XVII, 2)
with a dropping mercury electrode, an agar-KCl bridge connected to a SCE
through saturated potassium chloride solution contained in a 10-cm^ beaker, a
nitrogen gas inlet, and a magnetic stirrer. Charge the flask with 25.0 cm^ of the
iodide solution, add 25 cm^ 0. IM-nitric acid, and 2.5 cm^ warm 1, percent gelatin
solution. Connect the dropping mercury electrode to the negative terminal of a
polarograph and the positive terminal to the SCE. Set the applied potential at
zero and adjust the zero of the galvanometer at the, centre of the scale. Pass
nitrogen through the solution for at least 5 minutes whilst stirring magnetically.
Run in the mercury(II) nitrate solution from a semi-micro burette and take
readings of the galvanometer at 0.10-cm^ intervals. The end point corresponds
to zero current, ■ but continue the titration beyond this point to obtain the
cathodic current due to excess of mercury(II) nitrate. Plot galvanometer readings
against volume of mercury(II) nitrate solution, and evaluate the exact end point
from the graph.

The end point may be checked by potentiometric titration. Calculate the
concentration of the mercury(II) nitrate solution from the known concentration
of the potassium iodide solution ; alternatively, assume that the former is 0.0 IM
and calculate the molarity of the latter. •

Note. A standard solution of mercury(II) nitrate may be prepared by
dissolving a weighed amount of twice-distilled mercury in nitric acid, heating the
solution to expel oxides of nitrogen, and then diluting with distilled water to the
desired volume. This solution may be used for the determination of iodide.

XVn, 9. DETERMINATION OF POTASSIUM WITH SODIUM TETRA-
PHENYLBORON (GRAPHITE INDICATING ELECTRODE).

The tetraphenylboron ion (TPB) gives two anodic voltammetric waves at
a graphite electrode in aqueous solution. This electroactivity forms the basis
for the direct amperometric titration of potassium via its precipitation as
potassium tetraphenylboron. The method is simple and rapid; it is not
necessary to filter off the precipitate.

The procedure is relatively free from interferences, tolerating the presence of
large amounts of chloride and other commonly encountered anions, such as
phosphate, sulphate, and acetate. The tetraphenylboron ion forms insoluble
salts with ammonium, Rb, Cs, T1(I), Ag, and Hg(II) ions ; a precipitate is also
produced with Hg(I). These constitute the major interferences to the method.
Strong oxidising agents should be absent. The test solution should contain at
least 0.2 mg of K per cm^ ; below this concentration precipitation proceeds very
slowly, and the time required for a single measurement becomes excessive.

The procedure may be applied to the direct determination of potassium in
silicates and other refractory substances after sulphuric-hydrofluoric acid
dissolution and fuming.

Reagents required, (i) Prepare a 3 per cent sodium tetraphenylboron
solution by dissolving about 3.0 g of the pure solid reagent, accurately weighed,
in lOOcm^ of conductivity water. The solution is slightly turbid; satisfactory
results are obtained without removal of the turbidity;

681

XVII, 10 QUANTITATIVE INORGANIC ANALYSIS

(ii) Prepare a O-SA/ solution of sodium acetate (using the A.R, solid) and add
acetic acid until the pH is 5.6 (pH meter).

(iii) Prepare a ca. 0.02A/ solution potassium chloride using an accurately
weighed amount of the A.R. solid ; also a ca. 0.01 A/-potassium sulphate solution
employing the A.R. salt.

Apparatas. Prepare a saturated sodium chloride -calomel reference
electrode using A.R. sodium chloride : allow it to stand for 2-3 days before use. A
spectroscopic graphite elect rode about 1 0 cm long and 1 0 mm in diameter is used
as indicator electrode.

Use a potentiometer together with a sensitive galvanometer.

The titration vessel may be a three-necked tlask (see Fig. XVII, 2, A) of
lOOcnr* capacity. Insert liie arm of the saturated sodium chloride-calomel
electrode and the graphite electrode into the two side necks and a 5- or lO-cra^
semi-micro burette into the central neck. Connect the graphite electrode to the
potentiometer with the aid of an alligator dip. Stir the solution using a magnetic
stirrer, and maintain the same speed of stirring during all the determinations.
Alternatively, stir the solution with a glas.s stirrer at a constant speed ofabout600
r.p.m. : this will necessitate the use of a four-necked llask.

Procedure. Charge tile titration llask with 25 cm^ of the acetate bulTer
solution, introduce 25.0cnr' of the standard 0.02,l/-poiassium chloride, and add
sullicient water to ensure that at least I cm of the graphite electrode is immer-sed in

the solution. Apply a potential of +0.55 volt
to the graphite electrode. .Stir, and add the
sodium tetraphenylboron reagent from the
semi-micro burette (about 0.5-cm^ incremeiiLs
before the end point and 0.05- 0.1 0-cm^
increments after the end-(X)int). After each
addition, record the current as soon as it
becomes constant (1-3 minutes); after the
end point has been reached, the current is
Usually constant after 30 seconds. Determine
the eiul-poini by plotting galvanometer read-
ings against volume of liirant. as in Fig. XVII.
3. Calculate the litre of the reagent, i.c., rag
K c. 1.00 cm* Na TPB reagent.

Determine the pota.ssium content of the
0.0 1 A/-poiassium .sulphate solution and
compare the result obtained with that calculated from the weight of potassium
sulphate used. Alternatively, determine the potassium content of ‘unknown’
potassium chloride solutions (lU-^ocm-*) containing between 5 and 20 mg of
potassium. A new graphite electrode should be used for each determination.

/’

Titrations with the rotating platinum micro-clectrode

XVTl, 10. DISCUSSION AND ARPARATL'S. The dropping mercury
electrode cannot be used at markedly positive potentiaks (say, above about 0.4
volt w. SCE) because of the o.xidalion of the mercury'. By replacing the dropping
mercury electrode by an inert platinum electrode, it was iioped to e,vtend the
range ol polarogruphie work in the positive direction to the voltage approaching

SoUium lclrjphcn)lboratc sclulion. emt
Fig. XVII, 3

682

AMPEROMETRY XVII, 11

that at which oxygen is evolved, namely, 1.1 volts. The' attainment of a steady
diffusion current is slow with a stationary platinum electrode, but the difficulty
may be overcome by rotating the platinum electrode at constant speed: the
diffusion layer thickness is considerably reduced, thus increasing the sensitivity
and the rate of attainment of equilibrium. Difficulties,, however, arise in
obtaining reproducible values for the diffusion currents from day to day, and so /
the applications of the rotating platinum electrode in quantitative polarography

- I IaW

21-22 tnm

Mercury

4- Copper wire

Mercury

Fig.xvn,4

are limited. Nevertheless, it is suitable as an
indicator electrode in amperometric titrations. The
larger currents (about twenty times those at the
dropping mercury electrode) attained with the
rotating platinum electrode allow correspondingly
smaller currents to be measured without loss of
accuracy and thus very dilute solutions (up to
may be titrated. In order to obtain a linear
relation between current and amount of reagent
added, the speed of stirring must be kept constant
during the titration: a speed of about 600 rev-
olutions per minute* is generally suitable.

The construction of a simple rotating platinum
micro-electrode will be evident from Fig. XVII, 4.
The electrode is constructed from a standard

‘mercurjrsearrAbout 5 mm of platinum wire (0.5
mm diameter) protrudes from the wall ^f a length of
6->BHH-Blass tubing; the latter is benrat an angle

Platinum wile
•5- 6 mm long
0-5 mm diameter

approaching a right angle a short distance from the
lower end. Electrical connection is made to the
elecrode by a stout amalgamated copper wire
passing through the tubing to the mercury covering
the platinum wire seal ; the upper end of the copper
wire passes through a small hole blown in the stem
of the..stirrer and dips into mercury contained in the
‘mercury seal’. A wire from the latter is connected to
the source of applied voltage. The tubing forms the
stem of the electrode, which is rotated at a constant
speed of 600 r.p.m.

XVII, 11, DETERMINATION OF THIOSULPHATE WITH IODINE,
Dilute solutions of sodium thiosulphate (e.g., O.OOIM) may be titrated
with dilute iodine solutions (e.g., 0.005A/) at zero applied voltage. For. satis-
factory results, the thiosulphate solution slibuld be present in a supporting
electrolyte which is 0. 1 Af in potassium chloride and 0.004M in potassium iodide.
Under these conditions no diffusion current is detected until after the equivalence
point when excess of iodine is reduced at the electrode; a reversed L-type.of
titration graph is obtained.

Dilute solutions of iodine, e.g., O.OOOlAf, may be titrated similarly with

* The limiting
2QQ r.p.m.

current is proportional to the cube root of the number of revolutions per minute above

683

XVII, 12/13 QUANTITATIVE INORGANIC ANALYSIS

standard thiosulphate. The supporting electrolyte consists of 1.0,l/-liydrochloric
acid and 0.004A/-potassium iodide. No e.xtcrnal e.m.f. is required when a SCE is
cmploved as reference electrode.

Reagents required, (i) Prepare a cu. 0.001 ,l/-sodi urn thiosulphate solution
svhich is DAM with respect to potassium chloride and O.OOdM with respect to
potassium iodide.

(it) Prepare a standard 0.00.5d/-iodine solution in 0.0()4d/-potassium iodide.

Procedure, Place 25.0 crif' of the thiosulphate solution in the titration cell.
Set the applied voltage to zero with respect to the SCE. after connecting the
rotating platinum micro-electrode to the manual polarograph. Adjust the
sensitivity control of the galvanometer. Titrate with the standard 0.005d/-iodine
solution in the usual manner.

Plot the titration graph, evaluate the end [loint. and calculate the e.xact
concentration of the thiosulphate solution. As a check, repeat the titration using
freshly-prepared starch indicator, solution.

XVII, 12. DETER.MINATION OF ARSEMTE WITH STAND.VRD
IODINE SOLUTION. Dilute solutions of sodium arsenite le.g., 0.00053/1
may be titrated with standard iodine solution using a rotating platinum micro-
electrode and a SCE. The supporting electrolyte consi.si.s of 0.1.I/-potas.sium
chloride -r0.1,l/-sodium hydrugencarbonate + 0.004.t/-potassium iodide. A
reversed L-iype of titration graph results.

Reagents required, (i) Prepare a 0.0(K)5.1/-sodiuni arsenite solution which
is 0.1,1/ in potassium chloride and sodium hydrogencarbonate and 0.004,1/ in
potassium iodide.

(ii) Prepare standard 0.005.l/-iodinesolution.

Procedure. Pipette 25,0 cm^ of the .sodium arsenite .solution into the
titration Ihi.sk. Set the applied vi.'liage to zero to. .SCE ; adjust the sensitivity of the
galvanometer. Tiir.ite with the standard 0.005,l/-iodine in the usual manner.

Plot the titration graph, evaluate the end point, and calculate the
concentration of the arsenite solution. Check the end point with starch indicator.

XVII, 13. DETERMIN.VnON OF ANTI.MONY WITH STANDARD
POTASSIUM BROMATE SOLUTION. Dilute solutions of trivalent
antimony and arsenic ten. O.tXXlS.l/l may he titrated with standard O.OLY-
potassium bromate m a supporting electrolyte of .l/-hydrochlorjc acid
containing 0.05,l/-potassium bromide. The two electrodes arc a rotating
platinum micro-electrode and a SCE: the former is polarized to -t-0.2 volt. A
reversed L-type of titration graph is obtained.

Reagents required, (i) Prepare a 0.005,1/ solution of A.R. potassium
tintimony! tartrate by dissolving 1,625 g of the A.R. solid in I dm^ of distilled
water. Dilute 25.0cnr^ of tliis solution to 250cm^ with l.l/-hydrochIoric acid
which is 0.05.1/ in potassium bromide.

(ii) Prepare a standard O.OLV-potassium bromate solution from the A.R.
solid.

Procedure. Pipette 25.0 cm^ of the antimony solution into the titration
cell. Set tlie applied voltage at 0.2 volt v.v. SCE, and adjust the sensitivity control
of the galvanometer. Titrate in the usual manner, and calculate the concentration
of the antimony solution.

6S4

AMPEROMETRY XVII, 14/15

Biamperometric titrations

XVII, 14. GENERAL DISCUSSION. The titrations so far discussed in this
chapter have been concerned with the use of a reference electrode (usually SCE),
in conjunction with a polarised electrode (dropping mercury electrode or
rotating platinum micro-electrode). Titrations may also be performed in a
uniformly stirred solution, by using two small but similar platinum electrodes- to
which a small e.m.f. ( 1-100 millivolts) is applied: the end point is .usually shown
by either the disappearance or the appearance of a current flowing between the
two electrodes. For the method to be applicable the only requirement is that a
reversible oxidation-reduction system be present either before or after the end

point.

A simple apparatus suitable for this procedure is shown in Fig. XVII, 5. B is a
3-volt torch battery or 2-volt accumulator, M is a micro-ammeter, R is a 500-
ohm, 0.5-watt radio potentiometer, and EE are platinum electrodes. The

potentiometer is set so that there is a potential droo of
B,, about 80-100 millivolts across the electrodes.

In a titration with two in dicator electrodes and when

reactant involves a reversible'" system C^^g’,

\ Ij -)- 2 e 21 “), an appreciable current flows through the

V cell. The amountdfdxidisedTorm reduced at the cathode

j is^equal to that formed by oxidation of the reduced form

\ ^ at the anode. Both electrodes are depolarised until the

dE flE oxidised component or the reduced component of

the system has been consumed by a titrant. Aiisj::^.the

end, po int, only one electrode remains depolarised if the
s ^ s tiVantj(e.g., thiosulphate ion, 2 S 203 _^“j;;^ 8406 ^ 3 - 2 eJ
1 1 1 1 d^oes^not inyplye^a^rg^Yerjible systej^^ thus flows
5 ^ until the 'ehli point; at or Stef the endpoint the current is

. ^ ^ J zero or virtually zero. In the determination of iodine by

Fig. XVn, 5 titration with thiosulphate a rapid decrease in current is

observed in the neighbourhood of the end point and this
has led to ths-name-hlead-stoD end-point’. The complen^dntary-type of end point,
w^ch resembles' a reversed L-typ^*^ amperometric graph is probably ‘/more
desirable in pre^ctice, and is obtained in the titration of an irreversible c 6 up)e (say, y
thiosulphate) jjy a reversible couple ^(say, iodine) : the current is very low before/
the end point) and a very rapid increase in current Sgnals the end point. When
both system^'are reversiWe (e.g., irpn(II) ions with .cerium(IV) or permanganate
ions; applied potential 100 millivolts), the currentiis zero or, close to zero at the
equivalence point and a V-shaped titration graph results. , , ,

XVn, 15. TITRATION OF TfflOSULPHATE WITH IODINE (‘DEAD-
STOP END POINT’). Reagents required. Prepare a ca. O.OOlM-sodium
thiosulphate solution and also a standard 0.005M-iodine solution.

Procedure. Pipette 25.0 cm^ of the thiosulphate solution into the titration
cell e.g., a 150-cm^ Pyrex beaker. Insert two similar platinum wire or foil
electrodes* into the cell and connect to a manual polarograph or to the apparatus

* A length of 6-7 mm of platinum wire of 0.5 mm diameter sealed into a glass tube is satisfactory;
electncal connection is made by means of a copper wire dipping into a little mercury in contact with
the platinum wire. .

685

XVII, 16 QUANTITATIVE INORGANIC ANALYSIS

of Fig. XVII, 5. Apply O.IO volt across the electrodes. Adjust the sensitivity of the
‘spot’ galvanometer to obtain full-scale deflection for a current of 10-25
milliamperes. Stir the solution with a magnetic stirrer. Add the iodine solution
from a 5-cm^ semi-niicro burette slowly in the usual manner and read the current
(galvanometer deflection) after each addition of the titrant. When the current
begins to increase, stop the addition ; then add the titrant by small increments of
0.05 or 0.10 cm^ Plot the titration graph, evaluate the end point, and calculate
the concentration of the thiosulphate solution. It will be found that the current is
fairly constant until the end point is approached and increases rapidly beyond.

XVn, 16. DETERMINATION OF NITRATE. Discussion. ‘Dead-stop’
end point titrimetry may be apphed to the determination of nitrate ion by
titration with ammonium iron(II) sulphate solution in a strong sulphuric acid
medium:

4FeS0.,-l-2HN03 + 2H2S0., = 2 Fe 2 (S 04)3 + N 203 4-3H20

Two platinum electrodes are immersed in sulphuric acid of suitable strength
containing the nitrate ion to be determined and a potential of about 1 00 millivolts
is applied. Upon titration with 0.4M-ammonium iron(II) sulphate solution there
is an initial rise in current followed by a gradual fall, with a marked increase at the
end point: the latter is easily determined from a plot of galvanometer reading
against volume of iron solution added. The concentration of water should not be
allowed to rise above 25 per cent (w/w). The temperature of the solution should
not exceed 40 °C.

Reagents. Sulphuric acid, ca. 25 per cent v/v (Solution A). Add cautiously
250 cm^ of concentrated sulphuric acid to 750 cm^ of water, cool, and dilute to 1
litre. (Take care with this addition.)

Ammomum iron(II) sulphate solution, ca. 0.4M. Dissolve about 15.6g,
accurately weighed, of A.R. ammonium iron(II) sulphate in lOOcm^ of Solution
A.

Potassium nitrate solution, ca. O.iM. Dissolve about 3.0 g, accurately weighed,
of A.R. potassium nitrate in a small volume of Solution A and dilute to 100 cm^
in a graduated flask with concentrated sulphuric acid.

Procedure. Fit up the apparatus as follows. Into a 100-cm^ four-necked
Pyrex flat-bottom flask containing a polythene-covered stirring bar, insert two
platinum wire electrodes (0.5 mm diameter; held in position by corks) and a
thermometer respectively into the three side necks : insert the tip of a semi-micro
burette and a nitrogen inlet tube into the central neck. Place the flask in a beaker
charged with an ice-water mixture and clamp the flask in position : mount the
beaker on a magnetic stirrer. Pipette lO.Ocm^ of the potassium nitrate solution
into the flask, add 40 cm^ of concentrated sulphuric acid, and mix well with the
aid of the magnetic stirrer. Apply a polarisation voltage of about 100 millivolts:
use a galvanometer with adjustable sensitivity control to measure the current.
Titrate with the ammonium iron(II) sulphate solution while stirring vigorously:
adjust the galvanometer sensitivity to about ■^. The galvanometer reading will
decrease slightly as the end point is approached (indicated by the fading of the
pinkish-brown colour of the solution) and will increase steadily beyond the end-
point. The temperature of the solution has some influence upon the
galvanometer deflection, and so readings should preferably be taken when the
solution temperature is about 20 °C.

686

AMPEROMETRY XVII, 17

Determine the end point from the plot of galvanometer deflection against
volume of iron reagent. Calculate the weight of nitrate ion equivalent to 1 .0 cm^
of the 0.4M iron solution. ;

When dealing with small amounts of nitrate ion it is advisable to pass a current
of pure nitrogen through the solution before commencing the titration, and to
maintain an atmosphere of nitrogen in the flask throughout the titration.

If chloride is present, saturated aqueous silver acetate solution should be
added in amount slightly more than the calculated quantity prior to the addition
of concentrated sulphuric acid. The procedure may be applied to the routine
analysis of mixtures of nitric and sulphuric acids, and to the determination of
nitrogen in esters such as nitroglycerine and nitrocellulose;. the latter are easily
hydrolysed by strong sulphuric acid after dispersal in glacial acetic acid.

XVn, 17. DETERMINATION OF WATER WITH THE KARL FISCHER
REAGENT. For the determination of small amounts of water, Karl Fischer
(1935) proposed a reagent prepared by the action of sulphur dioxide upon a
solution of iodine in a mixture of anhydrous pyridine and anhydrous methanol.
Water reacts with this reagent in a two-stage process in which one molecule of
iodine disappears for each molecule of water present :

SO,

SCsHsN+Ii + SOi-hHjO = 2C5H5NH+r +C5H5N

(i)

SO,

C5H5N

•O

-hCHaOH ^CjHsN

OSO2OCH3

(ii)

The end point of the reaction is conveniently determined electrometrically
using the dead-stop end point procedure. If a small e.m.f. is applied across two
platinum electrodes immersed in the reaction mixture a current will flow as long
as free iodine is present, to remove hydrogen and depolarise the cathode. When
the last trace of iodine has reacted the current will decrease to zero or very close to
zero. Conversely, the technique may be combined with a direct titration of the
sample with the Karl Fischer reagent : here the current in the electrode circuit
suddenly increases at the first appearance of unused iodine in the solution.

The original Karl Fischer reagent prepared with an excess of methanol was
somewhat unstable and required frequent standardisation. It was found that the
stability was improved by replacing the methanol by 2-methoxyethanol (methyl
cellosolve), and a satisfactory reagent may be prepared by dissolving resublimed
iodine (133g) in pure anhydrous pyridine (425 cm contained in a dry, glass-
stoppered bottle, and then adding 2-methoxyethanol (425 cm^). With the bottle
cooled in an ice bath, anhydrous liquid sulphur dioxide (70 cm^) is added in small
portions from a graduated cylinder which is kept in an ice-salt bath. Usually, it is
not worth the trouble of preparing the reagent which may be purchased from the
normal suppliers of laboratory chemicals.

The reagent, whether purchased or prepared in the laboratory must be
standardised, and this may be done with pure disodium tartrate dihydrate which
contains 15.66 per cent water, or more commonly, by means of a solution of

687

XVII, 18 QUANTITATIVE INORGANIC ANALYSIS

water in methanol. This solution is prepared as follows. Fill a dry 1 dm^, glass-
stoppered graduated flask to within lOOcm^ of the mark with anhydrous
methanol ( < 0. 1 per cent of water), and place it in a thermostatically-controlled
water bath at 25 °C, together with a small flask containing about 200 cm^ of the
same methanol. Weigh out accurately about 15 g of distilled water into the dm^
flask and, after the contents have acquired the temperature of the water bath,
adjust the volume to the mark with methanol from the smaller flask.

The Karl Fischer procedure is best carried out with a commercial apparatus,
which may be purchased, with slight modifications, from many of the leading
laboratory supply houses. Basically, the instruihent will carry two burettes, one
for the reagent and the other for the standard solution of water in methanol.
Each burette is attached to a reservoir which may hold up to 1 litre of liquid, and
a series of guard tubes containing dessicant to prevent the ingress of atmospheric
moisture are provided. The titration vessel is fitted with an air-tight cover, and is
provided with a pair of bright platinum electrodes connected to a micro-
ammeter; provision is made for stirring the contents of the vessel by means of a
magnetic stirrer. The scale of the micro-ammeter is often marked with ‘Excess
reagent’ and ‘Excess water’ signs. The usual experimental procedure is to add a
slight excess of the reagent so that all the water in the sample under test is reacted,
and then the excess Fischer reagent is back-titrated with the standard water-in-
methanol solution.

The method is clearly confined to those cases where the test substance does not
react with either of the components of the reagent, nor with the hydrogen iodide
which is formed during the reaction with water: the following compounds
interfere in the Karl Fischer titration.

(i) Oxidising agents, such as chromates, dichromates, copper{II) and
iron(III) salts, higher oxides, and peroxides.

Mn02-l-4C5H5NH^-t-2r = Mn^-" +4C5HjN-(-l2 + 2H,0

(ii) Reducing agents, such as thiosulphates, tin(Il) salts and sulphides.

(iii) Compounds which can be regarded as forming water with the
components of the Karl Fischer reagent, for example :

(a) basic oxides—

ZnO + 2C5H5NH-" = Zn^^ -f 2C5H5N-t-H20;

(b) salts of weak oxy-acids—

NaHCOj-j-CjHsNH^ = Na+ -t-HjO-t-COj-bCsHsN

XVII, 18, DETERMINATION OF THE WATER CONTENT OF A SALT
HYDRATE. The Karl Fischer procedure may be applied to the determination
of water present in hydrated salts or which is absorbed on the surface of solids.
The procedure, where applicable, is more rapid and direct than the commonly
used drying process. A sample of the finely powdered solid, containing 5-10
millimols (90-180 mg) of water, is dissolved or suspended in 25 cm^ of dry
methanol in a 250-cm^ glass-stoppered graduated flask. The mixture is titrated
with standard Karl Fischer reagent to the usual electrometric end point. An end
point stable for 15 seconds usually indicates complete reaction. If the initial
titration is incomplete the mixture may be titrated at 10-second intervals until a
suitable end point is obtained. The water content of the methanol solution is

688

AMPEROMETRY XVII, 18

determined by a separate titration of an equal volume, and the titre of the sample
is reduced by this amount. The corrected titre is equivalent to the available water
in the sample.

Standardisation of the Karl Fischer reagent. By means of a standard solution
of water in methanol. Dry the beaker, stirrer, and electrode system using
acetone and a stream of dry air. Rapidly add 2-3 drops water from a weighing
bottle (fitted with a stopper and small dropper) to the beaker, and immediately fit
the beaker in position on the apparatus. Add the Karl Fischer reagent in 1-cm^
portions at about two-second intervals. Switch on the stirrer after a few cm^ have
been added, and continue .the titration • until a permanent iodine colour is
obtained. The meter needle will now swing over to ‘Excess reagent’. Back-titrate
the excess of Fischer reagent with the standard water-in-methanol solution at a
rate of about 1 drop per second until the meter needle begins to oscillate;
continue the titration until one drop causes a large deflection and the needle reads
‘Excess water’. Stirring must not be vigorous, and should be maintained at a
steady rate throughout the titration. Record the volumes of reactants added, and
also the exact weight of water used.

Calculate the strength of the Fischer reagent in terms of milligrams of water
percm^ of solution from both results. A useful check on the standard solution of
water in methanol is thus available.

Reproducible standardisation figures are sometimes difficult to obtain because
of variation in the amount of adsorbed water present in the apparatus. The
following modified procedure may be used. Transfer 20.0 cm^ anhydrous
methanol to the titration beaker, stir, and add the Karl Fischer reagent until
about 1.0 cm^ excess is present. Now titrate to the end point with the standard
water-in-methanol solution. Introduce 3 drops of water as rapidly as possible
through the side arm of the beaker, titrate with the Karl Fischer reagent until a
permanent iodine colour is obtained and the needle of the meter is at ‘Excess
reagent’, add 1 .0 cm^ more of the reagent. Titrate the excess of Fischer reagent
with the standard water-in-methanol solution. Run in a further 10.0 cm^ of the
water-in-methanol solution, titrate with the Karl Fischer reagent until a known
excess is present, and back-titrate the excess with the water-in-methanol solution.
Several determinations of the strength of the Karl Fischer reagent can thus be
made.

By disodium tartrate dihydrate. Place 25.0 cm^ absolute methanol in the
titration vessel and titrate with the Karl Fischer reagent. Add 0.5-0.6 g pure
.odium tartrate dihydrate (15.66 per cent water), accurately weighed, stir, and
titrate again with the Karl Fischer reagent. The salt dissolves completely before

the titration is completed.

Calculate the mg of water equivalent to 1 cm^ of the Karl Fischer reagent from
the formula:

mg of H,0 per cm^ = mg of sample x 0.1566
cm^ of reagent

Analysis of the hydrate. To determine the water content of the hydrate
ISO lum acetate is a satisfactory material for practising the technique), proceed
Fi 20 cm^ of anhydrous methanol in the titration vessel of the Karl

titrat^^ add a slight excess of the Karl Fischer reagent and then back

pres standard water-methanol mixture; this will remove any water

m the methanol and also water adsorbed on the surface of the vessel.

689

XVn, 19 QUANTITATIVE INORGANIC ANALYSIS

Immediately add to the methanol about 0.2 g of crystallised sodium acetate
which has been previously placed in a weighing bottle, stir the solution, add Karl
Fischer reagent until a slight excess is present, and then back titrate with the
water in methanol solution. Finally, reweigh the weighing bottle and calculate
the water content of the salt.

It may be noted that in some modern Karl Fischer titrators (e.g., the ‘Aquatest
IF marketed by the Photovolt Corporation) a reagent is used which is deficient in
iodine, and then in each determination the requisite amount of iodine is
generated electrolytically ; i.e., the determination is made coulometrically. This
procedure eliminates many of the problems associated with the instability of the
normal Karl Fischer reagent, and obviates the necessity for standardisation of
the reagent, with the result that determinations can be carried out with great
rapidity.

XVn, 19. Selected bibliography

1. P. Delahay (1954). New Instrumental Methods in Electrochemistry. New York;
Interscience Publishers.

2. G. Chariot and D. Bezier (1954). Methodes Electrochimiques d' Analyse. Paris;
Masson et Cie.

3. G. Chariot and D. Bezier (1957). Translated by R. C. Murray. Quantitative Inorganic
Analysis. Chapter XXV. ‘Amperometry’. London ; Methuen and Co.

4. K. G. Stone and H. G. Scholten (1952). The Dead-Stop End Point’, Analytical
Chemistry, 24,671.

5. L. M. Kolthoff (1954). ‘Relations between Voltammetry and Potentiometric and
Amperometric Titrations’, Analytical Chemistry, 26, 1 685.

6. P. Delahay (1955). ‘Voltammetry at Constant Current’, Analytical Chemistry, 27,
478.

7. D. L. Smith, D. R. Jamieson, and P. J. Elving (1960). ‘Direct Titration of Potassium
with Tetraphenylborate. Amperometric Equivalence-Point Detection’, Analytical
Chemistry, 32, 1253.

8. H. A. Laitinen. ‘Amperometric Titrations’, Analytical Chemistry, 1956, 28, 666; 1958,
30, 657; 1960,32, 180R.

9. J. Mitchell and D. M. Smith (1948). Aquametry : Application of the Karl Fischer
Reagent to Quantitative Analysis Involving IVater. New York; Interscience.

10. L. Meites (1963). Handbook of Analytical Chemistry. New York; McGraw-Hill.

11. J. T. Stock (1974). Amperometric Titrations (review article). Anal. Chem., 46, IR.

12. J. T. Stock (1965). ‘Amperometric Titrations’. New York; Interscience.

690

SPECTROANALYTICAL
PARTF METHODS

691

COLORin/IETRV

CHAPTER XVIII AMD SPECTROPHOTOIWETRY

XVni, 1. GENERAL DISCUSSION. The variation of the colour of a system
with change in concentration of some component forms the basis of what the
chemist commonly terms colorimetric analysis. The colour is usually due to the
formation of a coloured compound by the addition of an appropriate reagent, or
it may be inherent in the desired constituent itself. The intensity of the colour may
then be compared with that obtained by treating a known amount of the
substance in the same manner.

Colorimetry is concerned with the determination of the concentration of a
substance by measurement of the relative absorption of light with respect to a
known concentration of the substance. In visual colorimetry, natural or artificial
white light is generally used as a light source, and determinations are usually
made with a simple instrument termed a colorimeter or colour comparator. When
the eye is replaced by a photoelectric cell (thus largely eliminating the errors due
to the personal characteristics of each observer) the instrument is termed a photo-
electric colorimeter. The latter is usually employed with light contained within a
comparatively narrow range of wavelengths furnished by passing white light
through filters, i.e., materials in the form of plates of coloured glass, gelatin, etc.,
transmitting only a limited spectral region: the name filter photometer is
sometimes applied to such an instrument.

In spectrophotometric analysis a source of radiation is used that extends into
the ultraviolet region of the spectrum. From this, definite wavelengths of
radiation are chosen possessing a bandwidth of less than 1 nm. This process
necessitates the use of a more complicated and consequently more expensive
instrument. The instrument employed for this purpose is a spectrophotometer,
and as its name implies, is really two instruments in one cabinet — a spectrometer
and a photometer.

An optical spectrometer is an instrument possessing an optical system which
can produce dispersion of incident electromagnetic radiation, and with which
measurements can be made of the quantity of transmitted radiation at selected
wavelengths of the spectral range. A photometer is a device for measuring the
intensity of transmitted radiation or a function of this quantity. When combined
m the spectrophotometer the spectrometer and photometer are employed
conjointly to produce a signal corresponding to the difference between the
ransmitted radiation of a reference material and that of a sample at selected
wavelengths.

The chief advantage of colorimetric and spectrophotometric methods is that

693

XVIII, 1 QUANTITATIVE INORGANIC ANALYSIS

they provide a simple means for determining minute quantities of substances. The
upper limit of colorimetric methods is, in general, the determination of
constituents which are present in quantities of less than 1 or 2 per cent. The
development of inexpensive photoelectric colorimeters has placed this branch of
instrumental chemical analysis within the means of even the smallest teaching
institution.

In this chapter we are concerned with analytical methods that are based upon
the absorption of electromagnetic radiation. Light consists of radiation to which
the human eye is sensitive, waves of different wavelengths giving rise to light of
different colours, while a mixture of light of these wavelengths constitutes white
light. White light covers the entire visible spectrum 400-760 nm. The
approximate wavelength ranges of colours are given in Table XVIII, 1.

Table XVIII, 1. Approximate wavelengths of colours

Ultraviolet

<400 nm

Yellow

570-590 nm

Violet

400-450 nm

Orange

590-620 nm

Blue

450-500 nm

Red

620-760 nm

Green

500-570 nm

Infrared

>760 nm

Wavelength
(metres )

-U-

1 Picomcire -12 -
-II-

I Angstrom -10-
I N^iocncifc - 9

I Micrometre

Frequency

The visual perception of colour arises from the selective absorption of certain
wavelengths of incident light by the coloured object. The other wavelengths are
either reflected or transmitted, according to the nature of the object, and are

perceived by the eye as the colour of
the object. If a solid opaque object
appears white, all wavelengths are
reflected equally; if the object ap-
pears black, very little light of any
wavelength is reflected; if it appears
blue, the wavelengths that give the
blue stimulus are reflected, etc.

It must be emphasised that the
range of electromagnetic radiation
extends considerably beyond the
visible region. The approximate
limits of wavelength and frequency
for the various types of radiation,
including the frequency range of
sound waves, are shown in Fig.
XVIII, 1 (not drawn to scale); this
may be regarded as an electromag-
netic spectrum. It will be seen that y-
rays and X-rays have very short
wavelengths, while ultraviolet, vis-
ible, infrared and radio waves have
progressively longer wavelengths.
For colorimetry and spectro-
photometry, the visible region and

1 MiDimclre

1 Metre

I Kilometre

"T-

2M I
in I \Ai

Ultraviolet

-r

T-r-i

300

250

1600 ' 1400 1200 1000
50000 40000 30000

Infrared

“T ,

400 500600750 1500
800 600 400 200

20000 10000

Fig.XVm,!

Wavelength
Frequency

Wave number jjjg adjacent ultraviolet region are of
major importance.

694

COLORIMETRY AND SPECTROPHOTOMETRY XVUI, 2

Electromagnetic waves are usually , described in terms of (a) wavelength X
(distance between the peaks of waves in cm, unless otherwise specified), {b)
wavenumber v ' (number of waves per cm), and (c) the frequency v (nuniber of
waves per second). The three quantities are related as follows : , .

1 , Frequency

Wavelength

= Wavenumber =

Velocity of light

1_ _v

The units in common use are :

1 Angstrom unit = 1 A = 10“ metre = 10“® cm

0 _

1 Nanometre = 1 nm = 10 A = 10“ cm
1 Micrometre = l^m = 10“^ A = 10"*^ cm
Velocity of light = c = 2.99793 x 10® ms“‘

Wavenumber v = 1/A waves per cm

Frequency v = c/A » 3 x 10^°/A waves per second.

To comply completely with SI units these functions should be calculated using
the metre as the basic unit. It is, however, still common practice to use centimetres
for this purpose.

XVin, 2. THEORY OF SPECTROPHOTOMETRY* AND COLORI-
METRY. When light (monochromatic or heterogeneous) falls upon a
homogeneous medium, a portion of the incident light is reflected, a portion is
absorbed within the medium, and the remainder is transmitted. If the intensity of
the incident light is expressed by Iq, that of the absorbed light by I a, that of the
transmitted light by and that of the reflected light by 7^, then:

^o = 7a + 7t+7r

For air-glass interfaces consequent upon the use of glass cells, it may be stated
that about 4 per cent of the incident light is reflected. I, is usually eliminated by
the use of a control, such as a comparison cell, hence:

h-Ia+h ( 1 )

Credit for investigating the change of absorption of light with the thickness of
the medium is frequently given to Lambert (Ref. 1), although he really extended
concepts originally developed by Bouguer (Ref. 2). Beer (Ref. 3) later applied
similar experiments to solutions of different concentrations and published his
results just prior to those of Bernard (Ref. 4). This very confusing story has been
explained by Malinin and Yoe (Ref 5). The two separate laws governing
a sorption are usually known as Lambert’s Law and Beer’s Law. In the combined
orm (Ref 6) they are referred to as the Beer-Lambert Law.

Lambert’s law. This law states that when monochromatic light passes

pec rophotometry proper is mainly concerned with the following regions of the spectrum:
with th° visible 400-760 nm; and infrared, 0.76-15 pm. Colorimetry is concerned

wciKi ^ region of the spectrum, fn this chapter our attention will be confined largely to the
°'e =‘"‘1 near ultraviolet region of the spectrum.

.695

XVIII, 2 QUANTITATIVE INORGANIC ANALYSIS

through a transparent medium, the rate of decrease in intensity with the thickness
of the medium is proportional to the intensity of the light. This is equivalent to
stating that the intensity of the emitted light decreases exponentially as the
thickness of the absorbing medium increases arithmetically, or that any layer of
given thickness of the medium absorbs the same fraction of the light incident
upon it. We may express the law by the differential equation :

where / is the intensity of the incident light of wavelength )., / is the thickness of
the medium, and k is a proportionality factor. Integrating (2) and putting / = /^
when I = 0, we obtain:

In^ = kl

or, stated in other terms,

= ( 3 )

where /q is the intensity of the incident light falling upon an absorbing medium of
thickness /, I, is the intensity of the transmitted light, and k is a constant for the
wavelength and the absorbing medium used. By changing from natural to
common logarithms we obtain:

/, = /o-10-°‘^^‘*“' = /o-10-'''' (4)

where K = /c/2.3026 and is usually termed the absorption coefficient. The
absorption coefficient is generally defined as the reciprocal of the thickness (/ cm)
required to reduce the light to of its intensity. This follows from equation (4),
since:

IJlo = 0.1 = 10-'^' or iC/ = 1 and K = l/l

The ratio IJIg is the fraction of the incident light transmitted by a thickness I of
the medium and is termed the transmittance T. Its reciprocal Ig/I, is the opacity,
and the absorbance A of the medium (formerly called the optical density D or
extinction E) is given by:

A = log!o/I, (5)

Thus a medium with absorbance 1 for a given wavelength transmits 10 per cent of
the incident light at the wavelength in question.

Beer’s law. We have thus far considered the light absorption and the light
transmission for monochromatic light as a function of the thickness of the
absorbing layer only. In quantitative analysis, however, we are mainly concerned
with solutions. Beer studied the effect of concentration of the coloured
constituent in solution upon the light transmission or absorption. He found the
same relation between transmission and concentration as Lambert had discovered
between transmission and thickness of the layer {equation (3)}, i.e., the intensity
of a beam of monochromatic light decreases exponentially as the concentration
of the absorbing substance increases arithmetically. This may be written in the
form:

I = ^0

696

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 2

where c is the concentration, and k' and K' are constants. Combining (4) and (5),
we have (Ref. 6):

= (7)

or logJo/i, = acl (8)

This is the fundamental equation of colorimetry and spectrophotometry, and is
often spoken of as the Beer-Lambert law. The value of a will clearly depend upon
the method of expression of the concentration. If c is expressed in mole dm“ ^ and
/ in centimetres then-a is given the symbol e and is called the molar absorption
coefficient or molar absorptivity (formerly the molar extinction coefficient).

The specific absorption (or extinction) coefficient E^- (sometimes termed
absorbancy index) may be defined as the absorption per unit thickness (path
length) and unit concentration.

Where the molecular weight of a substance is not definitely known, it is
obviously not possible to write down the molecular absorption coefficient, and in
such cases it is usual to write the unit of concentration as a superscript, and the
unit of length as a subscript.

Thus £j“4, 325 nm = 30

means that for the substance in question, at a wavelength of 325 nm, a solution of
length 1 cm, and concentration 1 per cent (1 per cent by weight of solute or 1 g of
solid per 100 cm^ of solution) log Iq/I, has a value of 30.

It will be apparent that there is a relationship between the Absorbance A, the
Transmittance T, and the molar absorption coefficient, since:

d = ec/ = log^ = logi= -logT (9)

The scales of spectrophotometers are often calibrated to read directly in
absorbances, and frequently also in percentage transmittance. It may be
mentioned that for colorimetric measurements Iq is usually understood as the
intensity of the light transmitted by the pure solvent, or the intensity of the light
entering the solution; /, is the intensity of the light emerging from the solution, or
transmitted by the solution. It will be noted that;

the absorption coefficient (or extinction coefficient) is the absorbance for unit
path length

K = Ajt or /, = Iq -

the specific absorption coefficient (or absorbancy index) is the absorbance per
unit path length and unit concentration

£, = A/d or /, =

the molar absorption coefficient (or molar extinction coefficient) is the specific
absorption coefficient for a concentration of 1 mole dm“ ^ and a path length
of 1 cm.

£ = A/d

Application of Beer’s law. Let us consider the case of two solutions of a
co oured substance with concentrations Ci and c,. These are placed in an
ns rument in which the thickness of the layers can be altered and measured
S' y, and which also allows a comparison of the transmitted light (e.g., a

697

XVIII, 2 QUANTITATIVE INORGANIC ANALYSIS

= kl (2)

where I is the intensity of the incident light of wavelength 2, / is the thickness of
the medium, and k is a proportionality factor. Integrating (2) and putting / = /^
when / = 0, we obtain:

ln^ = k/

or, stated in other terms,

= (3)

where /q is the intensity of the incident light falling upon an absorbing medium of
thickness /, /, is the intensity of the transmitted light, and k is a constant for the
wavelength and the absorbing medium used. By changing from natural to
common logarithms we obtain :

/, = = (4)

where K = k/2.3026 and is usually termed the absorption coefficient. The
absorption coefficient is generally defined as the reciprocal of the thickness (/ cm)
required to reduce the light to ^ of its intensity. This follows from equation (4),
since:

= 0.1 = 10-'"' or K/ = 1 and K = 1//

The ratio IJlg is the fraction of the incident light transmitted by a thickness I of
the medium and is termed the transmittance T. Its reciprocal /q//, is the opacity,
and the absorbance A of the medium (formerly called the optical density D or
extinction E) is given by:

A = logIo/l, (5)

Thus a medium with absorbance 1 for a given wavelength transmits 10 percent of
the incident light at the wavelength in question.

= J^.lQ-0.4343k'c^j^.lO-K-c ( 6 )

696

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 2

where c is the concentration, and k' and K' are constants. Combining (4) and (5),
we have (Ref. 6);

= (7)

or log/fl//, = flc/ (8)

This is the fundamental equation of colorimetry and spectrophotometry, and is
often spoken of as the Beer-Lambert law. The value of a will clearly depend upon
the method of expression of the concentration. If c is expressed in mole dm ~ ^ and
I in centimetres then- a is given the symbol e and is called the molar absorption
coefficient or molar absorptivity (formerly the molar extinction coefficient).

The specific absorption (or extinction) coefficient E, (sometimes termed
absorbancy index) may be defined as the absorption per unit thickness (path
length) and unit concentration.

Thus £ri 325nm = 30

It will be apparent that there is a relationship between the Absorbance A, the
Transmittance T, and the molar absorption coefficient, since;

A = ecl = log^ = log^ = -log T (9)

The scales of spectrophotometers are often calibrated to read directly in
absorbances, and frequently also in percentage transmittance. It may be
mentioned that for colorimetric measurements Iq is usually understood as the
intensity of the light transmitted by the pure solvent, or the intensity of the light
entering the solution; I, is the intensity of the light emerging from the solution, or
transmitted by the solution. It will be noted that:

the absorption coefficient (or extinction coefficient) is the absorbance for unit
path length

K = A/t or/, =

the specific absorption coefficient (or absorbancy index) is the absorbance per
unit path length and unit concentration

E, = A/cI or/, = /o-10-'^*‘''

the molar absorption coefficient (or molar extinction coefficient) is the specific
absorption coefficient for a concentration of 1 mole dm~^ and a path length
of 1 cm.

£ = A/cl

Application of Beer’s law. Let us consider the case of two solutions of a
CO cured substance with concentrations Cj and C 2 - These are placed in an
^ s rument in which the thickness of the layers can be altered and measured

SI y, and which also allows a comparison of the transmitted light (e.g., a

697

XVm, 2 QUANTITATIVE INORGANIC ANALYSIS

Duboscq colorimeter, Section XVIII, 6). When the two layers have the, same
colour intensity:

= /o • = /,, = lo ■ 10-^'^'=^ (10)

Here Ii and I 2 are the lengths of the columns of solutions with concentrations
and C 2 respectively when the system is optically balanced. Hence under these
conditions and when Beer’s law holds:

/jCj = / 2 C 2 (11)

A colorimeter can therefore be employed in a dual capacity: (a) to investigate the
validity of Beer’s law by varying c^ and C 2 and noting whether equation (11)
applies, and {b) for the determination of an unknown concentration of a
coloured solution by comparison with a solution of known concentration o,. It
must be emphasised that equation ( 1 1) is valid only if Beer’s law is obeyed over the
concentration range employed and the instrument has no optical defects.

When a spectrophotometer is used it is unnecessary to make comparison with
solutions of known concentration. With such an instrument the intensity of the
transmitted light or, better, the ratio IJl^ (the transmittance) is found directly at a
known thickness /. By varying / and c the validity of the Lambert-Beer law,
equation (6), can be tested and the value of c may be evaluated. When the latter is
known, the concentration of an unknown solution can be calculated from the
formula:

( 12 )

Attention is directed to the fact that the extinction coefficient e depends upon the
wavelength of the incident light, the temperature, and the solvent employed. In
general, it is best to work with light of wavelength approximating to that for
which the solution exhibits a maximum selective absorption (or minimum
selective transmittance): the maximum sensitivity is thus attained.

For matched cells (i.e., I constant) the Beer-Lambert law may be written:

1 ^0

ccclog —

ccclogy

or coc A (13)

Hence by plotting A |or lt>g~|, as ordinate, against concentration as abscissa, a

straight line will be obtained and this will pass through the point c = 0, ^4 = 0
(T = 100%). This calibration line may then be used to determine unknown
concentrations of solutions of the same material after measurement of
absorbances.

Deviation from Beer’s law. Beer’s law will generally hold over a wide range
of concentration if the structure of the coloured ion or of the coloured non-
electrolyte in the dissolved state does not change with concentration. Small
amounts of electrolytes, which do not react chemically with the coloured
components, do not usually affect the light absorption; large amounts of

698

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 3

electrolytes may result in a shift of the maximum absorption, and may also
change the value of the extinction coefficient. Discrepancies are usually found
when the coloured solute ionises, dissociates, or associates in solution, since the
nature of the species in solution will vary with the concentration. The law does
not hold when the coloured solute forms complexes, the composition of which
depends upon the concentration. Also discrepancies may occur when
monochromatic light is not used. The behaviour of a substance can always be
tested by plotting loglo/I,, or logT against the concentration: a straight line
passing through the origin indicates conformity to the law.

For solutions which do not follow Beer’s law, it is best to prepare a calibration
curve using a series of standards of known concentration. Instrumental readings
are plotted as ordinates against concentrations in, say, mg per 100 cm^ or 1000
cm^ as abscissae. For the most precise work each calibration curve should cover
the dilution range likely to be met with in the actual comparison.

XVm, 3. CLASSIFICATION OF METHODS OF ‘COLOUR’ MEASURE-
MENT OR COMPARISON. The basic principle of most colorimetric
measurements consists in comparing under well-defined conditions the colour
produced by the substance in unknown amount with the same colour produced
by a known amount of the material being determined. The quantitative
comparison of these two solutions may, in general, be carried out by one or more
of six methods. It is not essential to prepare a series of standards with the
spectrophotometer; the molar absorption coefficient can be calculated from one
measurement of the absorbance or transmittance of a standard solution, and the
unknown concentration can then ■ be computed with the aid of the molar
absorption coefficient and the observed value of the absorbance or transmittance
(cf. Section XVin, 2, equations (12) and (13)).

A. Standard series method. (Section XVHI, 4) The test solution
contained in a Nessler tube is diluted to a definite volume, thoroughly mixed, and
Its colour compared with a series of standards similarly prepared. The
concentration of the unknown is then, of course, equal to that of the known
solution whose colour it matches exactly. The accuracy of the method will
depend, inter alia, upon the concentrations of the standard series; the probable
error is of the order of ± 3 per cent, but may be as high as ± 8 per cent.

For convenience, artificial standards, e.g., Lovibond glasses, salt solutions such
as iron(III) chloride in aqueous hydrochloric acid (yellow), aqueous cobalt chlor-
ide (pink), aqueous copper sulphate (blue), and aqueous potassium dichromate
(orange) are sometimes used. It is essential to standardise the artificial standards
against known amounts of the substance being determined, the latter always
being treated under exactly similar conditions. The disadvantage of this method
IS that the spectral absorption curves of the test solutions and of the sub-standard
glasses or solutions may be far from identical; the error due to this cause is greatly
®agmfied in the case of observers suffering from partial colour blindness.

B- Duplication method. (Section XVIII, 5) A standard solution of the
component under determination is added to the reagent until the colour
pro need matches that of the unknown sample in the same volume of solution.

this method is less accurate than A.

ol ^ J^Bution method. The sample and standard solution are contained in
o ass tubes of the same diameter, and are observed horizontally through the tubes.

e more concentrated solution is diluted until the colours are identical in

699

XVIII, 2 QUANTITATIVE INORGANIC ANALYSIS

Duboscq colorimeter, Section XVIII, 6). When the two layers have the same
colour intensity:

1 ,^ = /o ■ 10-''^“' = = /o • (10)

Here /j and are the lengths of the columns of solutions with concentrations Cj
and c ‘2 respectively when the system is optically balanced. Hence under these
conditions and when Beer’s law holds:

/jC^ = / 2 C 2 (11)

A colorimeter can therefore be employed in a dual capacity : (a) to investigate the
validity of Beer’s law by varying c, and C 2 and noting whether equation (11)
applies, and (6) for the determination of an unknown concentration C 2 of a
coloured solution by comparison with a solution of known concentration Cj. It
must be emphasised that equation (1 1) is valid only if Beer’s law is obeyed over the
concentration range employed and the instrument has no optical defects.

When a spectrophotometer is used it is unnecessary to make comparison with
solutions of known concentration. With such an instrument the intensity of the
transmitted light or, better, the ratio IJIq (the transmittance) is found directly at a
known thickness 1. By varying / and c the validity of the Lambert-Beer law,
equation (6), can be tested and the value of c may be evaluated. When the latter is
known, the concentration of an unknown solution can be calculated from the
formula:

log fp/f,

( 12 )

Attention is directed to the fact that the extinction coefficient £ depends upon the
wavelength of the incident light, the temperature, and the solvent employed. In
general, it is best to work with light of wavelength approximating to that for
which the solution exhibits a maximum selective absorption (or minimum
selective transmittance): the maximum sensitivity is thus attained.

For matched cells (i.e., I constant) the Beer-Lambert law may be written:-

ccclog —

c oi log -4
T

or cccA (13)

Hence by plotting A |or log:p|, ns ordinate, against concentration as abscissa, a

straight line will be obtained and this will pass through the point c = 0, A = 0
(T = 100%). This calibration line may then be used to determine unknown
concentrations of solutions of the same material after measurement of
absorbances.

698

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 3

electrolytes may result in a shift of the maximum absorption, and may also
change the value of the extinction coefficient. Discrepancies are usually found
when the coloured solute ionises, dissociates, or associates in solution, since the
nature of the species in solution will vary with the concentration. The law does
not hold when the coloured solute forms complexes, the composition of which
depends upon the concentration. Also discrepancies may occur when
monochromatic light is not used. The behaviour of a substance can always be
tested by plotting loglo/I,, or logT against the concentration: a straight hne
passing through the origin indicates conformity to the law.

XVni, 3. CLASSIFICATION OF METHODS OF ‘COLOUR’ MEASURE-
MENT OR COMPARISON. The basic principle of most colorimetric
measurements consists in comparing under well-defined conditions the colour
produced by the substance in unknown amount with the same colour produced
by a known amount of the material being determined. The quantitative
comparison of these two solutions may, in general, be carried out by one or more
of six methods. It is not essential to prepare a series of standards with the
spectrophotometer; the molar absorption coefficient can be calculated from one
measurement of the absorbance or transmittance of a standard solution, and the
unknown concentration can then be computed with the aid of the molar
absorption coefficient and the observed value of the absorbance or transmittance
(cf. Section XVIII, 2, equations (12) and (13)).

A Standard series method. (Section XVIII, 4) The test solution
contained in a Nessler tube is diluted to a definite volume, thoroughly mixed, and
its colour compared with a series of standards similarly prepared. The
concentration of the unknown is then, of course, equal to that of the known
solution whose colour it matches exactly. The accuracy of the method will
depend, inter alia, upon the concentrations of the standard series; the probable
error is of the order of + 3 per cent, but may be as high as ± 8 per cent.

For convenience, artificial standards, e.g., Lovibond glasses, salt solutions such
as iron(III) chloride in aqueous hydrochloric acid (yellow), aqueous cobalt chlor-
ide (pink), aqueous copper sulphate (blue), and aqueous potassium dichromate
(orange) are sometimes used. It is essential to standardise the artificial standards
against known amounts of the substance being determined, the latter always
being treated under exactly similar conditions. The disadvantage of this method
is that the spectral absorption curves of the test solutions and of the sub-standard
glasses or solutions may be far from identical ; the error due to this cause is greatly
fflagnified in the case of observers suffering from partial colour blindness.

B- Duplication method. (Section XVIII, 5) A standard solution of the
component under determination is added to the reagent until the colour
pro uced matches that of the unknown sample in the same volume of solution.

bis method is less accurate than A.

] ^Mutton method. The sample and standard solution are contained in
g ass tubes of the same diameter, and are observed horizontally through the tubes.

c more concentrated solution is. diluted until the colours are identical in

699

XVni, 2 QUANTITATIVE INORGANIC ANALYSIS

Duboscq colorimeter, Section XVIII, 6). When the two layers have the same
colour intensity;

= lo ■ = /,, - lo ■ (10)

Here /j and are the lengths of the columns of solutions with concentrations Cy
and C 2 respectively when the system is optieally balanced. Hence under these
conditions and when Beer’s law holds;

lyCy = / 2 C 2 (11)

A colorimeter can therefore be employed in a dual capacity; (a) to investigate the
validity of Beer’s law by varying Cy and C 2 and noting whether equation (11)
applies, and (b) for the determination of an unknown concentration of- a
coloured solution by comparison with a solution of known concentration Cy. It
must be emphasised that equation (1 1) is valid only if Beer’s law is obeyed over the
concentration range employed and the instrument has no optical defects.

When a spectrophotometer is used it is unnecessary to make comparison with
solutions of known concentration. With such an instrument the intensity of the
transmitted light or, better, the ratio IJI^ (the transmittance) is found directly at a
known thickness /. By varying / and c the validity of the Lambert-Beer law,
equation (6), can be tested and the value of c may be evaluated. When the latter is
known, the concentration of an unknown solution can be calculated from the
formula;

log/o/f,

( 12 )

Attention is directed to the fact that the extinction coefficient c depends upon the
wavelength of the incident light, the temperature, and the solvent employed. In
general, it is best to work with light of wavelength approximating to that for
which the solution exhibits a maximum selective absorption (or minimum
selective transmittance); the maximum sensitivity is thus attained.

For matched cells (i.e., / constant) the Beer-Lambert law may be written;

coclog —

, 1
ccclog-

or coc A (13)

Hence by plotting A |or ordinate, against concentration as abscissa, a

straight line will be obtained and this will pass through the point c = 0, A — 0
(T = 100%). This calibration line may then be used to determine unknown
concentrations of solutions of the same material after measurement of
absorbances.

698

COLORIMETRY AND SPECTROPHOTOMETRY XVIU, 3

electrolytes may result, in a shift of the maximum absorption, and may also
change the value of the extinction coefficient. Discrepancies are usually found
when the coloured solute ionises, dissociates, or associates in solution, since the
nature of the species in solution will vary with the concentration. The law does
not hold when the coloured solute forms complexes, the composition of which
depends upon the concentration. Also discrepancies may occur- when
monochromatic light is not used. The behaviour of a substance can always be
tested by plotting log/o/fr. or log T against the concentration: a straight line
passing through the origin indicates conformity to the law.

XVm, 3. CLASSIFICATION OF METHODS OF ‘COLOUR’ MEASURE-
MENT OR COMPARISON. The basic principle of most colorimetric
measurements consists in comparing under well-defined conditions the colour
produced by the substance in unknown amount with the same colour produced
by a known amount of the materia! being determined. The quantitative
comparison of these two solutions may, in general, be carried out by one or more
of six methods. It is not essential to prepare a series of standards with the
spectrophotometer; the molar absorption coefficient can be calculated from one
measurement of the absorbance or transmittance of a standard solution, and the
unknown concentration can then be computed with the aid of the molar
absorption coefficient and the observed value of the absorbance or transmittance
(cf. Section XVHI, 2, equations (12) and (13)).

A. Standard series method. (Section XVHI, 4) The test solution
contained in a Nessler tube is diluted to a definite volume, thoroughly mixed, and
its colour compared with a series of standards similarly prepared. The
concentration of the unknown is then, of course, equal to that of the known
solution whose colour it matches exactly. The accuracy of the method will
depend, inter alia, upon the concentrations of the standard series; the probable
error is of the order of ± 3 per cent, but may be as high as ± 8 per cent.

For convenience, artificial standards, e.g., Lovibond glasses, salt solutions such
as iron(III) chloride in aqueous hydrochloric acid (yellow), aqueous cobalt chlor-
ide (pink), aqueous copper sulphate (blue), and aqueous potassium dichromate
(orange) are sometimes used. It is essential to standardise the artificial standards
against known amounts of the substance being determined, the latter always
being treated under exactly similar conditions. The disadvantage of this method
IS that the spectral absorption curves of the test solutions and of the sub-standard
g asses or solutions may be far from identical ; the error due to this cause is greatly
Magnified in the case of observers suffering from partial colour blindness.

B. Duplication method. (Section XVIII, 5) A standard solution of the
coniponent under determination is added to the reagent until the colour
pro need matches that of the unknown sample in the same volume of solution.

bis method is less accurate than A.

1 dilution method. The sample and standard solution are contained in

ass tubes of the same diameter, and are observed horizontally through the tubes.

e more concentrated solution is diluted until the colours are identical in

699

XVIII, 3 QUANTITATIVE INORGANIC ANALYSIS

intensity when observed horizontally through the same thickness of solution. The
relative concentrations of the original solutions are then proportional to the
heights of the matched solutions in the tubes. This is the least accurate method of
all, and will not be discussed further.

D. Balancing method. (Section XVIII, Q This method forms the basis of
all colorimeters of the plunger type, e.g., in the Duboscq colorimeter. The
comparison is made in two tubes, and the height of the liquid in one tube is
adjusted so that when both tubes are observed vertically the colour intensities in
the tubes are equal. The concentration in one of the tubes being known, that in the
other may be calculated from the respective lengths of the two columns of liquid
and the relation (eqn XVIII, 1 1):

It must be emphasised again that this simple proportionality holds only if Beer’s
law is applicable, and that the relation holds with greater exactness if a beam of
monochromatic light (obtained with the aid of a suitable colour filter) rather than
white light is employed. As a general rule, it is preferable that the solutions under
comparison should not differ greatly in concentration, and for the most accurate
work an empirically constructed calibration curve should be used. As usually
employed with white light, the accuracy obtainable with a Duboscq colorimeter
is of the order of ±7 per cent; the accuracy is increased appreciably if
monochromatic light (produced with colour filters) is employed.

E. Photoelectric photometer method. (Section XVIII, 7) In this method
the human eye is replaced by a suitable photoelectric cell; the latter is employed
to afford a direct measure of the light intensity, and hence of the absorption.
Instruments incorporating photoelectric cells measure the light absorption and
not the colour of the substance: for this reason the term ‘photoelectric
colorimeters’ is a misnomer; better names are photoelectric comparators,
photometers, or, best, absorptiometers.

Essentially most such instruments consist of a light source, a suitable light filter
to secure an approximation to monochromatic light (hence the name
photoelectric filter photometer), a glass cell for the solution, a photoelectric cell to
receive the radiation transmitted by the solution, and a measuring device to
determine the response of the photoelectric cell. The comparator is first
calibrated in terms of a series of solutions of known concentration, and the results
plotted in the form of a curve connecting concentrations and readings of the
measuring device employed. The concentration of the unknown solution is then
determined by noting the response of the cell and referring to the calibration
curve.

These instruments are available in a number of different forms incorporating
one or two photocells. With the one-cell type, the absorption of light by the
solution is usually measured directly by determining the current output of the
photoelectric cell in relation to the value obtained with the pure solvent. It is of
the utmost importance to use a light source of constant intensity, and if the photo
cells exhibit a ‘fatigue effect’ it is necessary to allow them to attain its
equilibrium current after each change of light intensity. The two-cell type of filter
photorneter is usually regarded as the more trustworthy (provided the electrical
circuit is appropriately designed) in that any fluctuation of the intensity of the
light source will affect both cells alike if they are matched for their spectral
response. Here the two photocells, illuminated by the same source of light, are

700

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 4

balanced against each other through a galvanometer; the test solution is placed
before one cell and the pure solvent before the other, and the current output
difference is measured.

F. Spectrophotometer method. (Section XVIH, 9) This is undoubtedly
the most accurate method for determining inter alia the concentration of
substances in solution, but the instruments are, of necessity, more expensive. A
spectrophotometer may be regarded as a refined filter photoelectric photometer
which permits the use of continuously variable and more nearly monochromatic
bands of light. The essential parts of a spectrophotometer are: (i) a source of
radiant energy, (ii) a monochromator, i.e., a device for isolating monochromatic
light or, more accurately expressed, narrow bands of radiant energy from the light
source, (iii) glass or silica cells for the solvent and for the solution under test, and
(iv) a device to receive or measure the beam or beams of radiant energy passing
through the solvent or solution.

In the following Sections it is proposed to discuss the more important of the
above methods in somewhat greater detail. For a more complete treatment the
reader is referred to the special treatises on the subject (see Selected Bibliography
at end of chapter).

XVni, 4. STANDARD SERIES METHOD. In this method colourless glass
tubes of uniform cross-section and with flat bottoms are usually employed. These
are termed Nessler tubes. The best variety have polished, flat bottoms. They are
made in either the ‘low’ form with a height of 175-200 mm and a diameter of 25-
32 mm (Fig. XVIII, 2) or as a ‘high’ form with a height of 300-375 mm and a
diameter of 21-24 mm. The solution of the substance being determined is made

Fig.XVin,2 Fig.XVin,3

up to a definite volume, and the colour is compared with that of a series of
standards prepared in the same way from known amounts of the component
eing determined. Fifty or 100 cm^ of the unknown and standard solutions are
P aced in Nessler tubes, and the solutions are viewed vertically through the length
0 the columns of the liquid. The concentration of the unknown is equal to that of
e staiidard having the same colour.* As a general rule, it will be found that the
CO our intensity of the unknown lies between two successive standards. Another

possible, to make a preliminary determination of the strength of the
a Ne adding from a burette a solution of the component in known concentration to

colou^ lube containing the reagents diluted with a suitable amount of water until the depth of
cental practically the same as that of an equal volume of the unknown solution also

rnn.-» T cylinder and standing at its side. A series of standards on either side of this

concentration IS then prepared.

701

XVIII, 4 QUANTITATIVE INORGANIC ANALYSIS

series of standards may then be prepared covering the latter range over smaller
concentration intervals. Thus, for example, in the determination of a particular
constituent the first series of standards might cover the range 0. 1 , 0.2, 0.4, 0.6, 0.8,
and 1.0 mg dm " and it is found that the colour of the unknown lies between 0.4
and 0.6 mg dm“^. The second series of standards may then be prepared
containing 0.40, 0.45, 0.50, 0.55, and 0.60 mg dra"^. Further comparison may
then show that the value lies between 0.45 and 0.50 mg dm~^, and for many
purposes this should be returned as 0.48 mg dm“^. If a more accurate value is
required, and provided the colour intensity of the solution and also the apparatus
employed will permit of finer comparison, another series of standards covering
the range of, say, 0.45, 0.475, and 0.50 mg dm~^ may be made up and the
unknown compared with these standards.

For the comparison of colours in Nessler tubes, the simplest apparatus consists
of a modified test-tube rack (Fig. XVIII, 3). It is constructed of wood, finished dull
black, and is provided with an inclined opal glass reflector or mirror, arranged to
reflect light up through the tubes. The Nessler tubes rest on a narrow ledge, and
do not come into contact with the reflector. The unknown and standards are
compared by placing them adjacent to each other and looking vertically down
through them.

This procedure serves as the basis for the colorimetric determination of pH by
employing a series of buffer solutions and suitable indicators. A series of appro-
priate buffer solutions is selected, differing successively in pH by about 0.2,
covering the pH range of the solutions under investigation; the range of the
buffer solutions required will be indicated by the preliminary pH determina-
tion. Equal volumes, say 10 cm^, of the buffer solutions differing successively
in pH by about 0.2 are placed in test-tubes of colourless glass and having
approximately the same dimensions, and a small equal quantity of a suitable
indicator for the particular pH range is added to each tube. A series of different
colours corresponding to the different pH values is thus obtained. An equal
volume (say 10 cm^) of the test solution is treated with an equal volume of
indicator to that used for the buffer solutions, and the resulting colour is
compared with that of the coloured standard buffer solutions. When a complete
match is found, the test solution and the corresponding buffer solution have the

same pH. Sometimes a complete match is not
obtained, but the colour of the test solution falls
between those of two successive standards, then
it is known that the pH value lies between those
of the two standards. Further buffer solutions
may then be prepared differing by 0.1 pH, if
desired, and pH value redetermined. As a general
rule, colorimetric methods cannot be relied upon
to give values of pH more accurate than to within
0.2 pH unit. For matching the colours, the buffer
solutions may be arranged in the holes of a
test-tube stand in order of pH: the test solution is
then moved from hole to hole until the best
colour match is obtained Special stands and
standards for making the comparision are
available commercially. The commercial stan-
dards, prepared from buffer solutions, are not

rr--

r^'

-Light-

Fig.XVm,4

702

COLORIMETRY AND SPECTROPHOTOMETRY XYHI, 4

permanent, and must be checked every six months.

For turbid or shghtly coloured solutions, the direct-comparisoii method given
above can no longer be applied. The interference due to the coloured substance
can be eliminated in a simple way by the following device, suggested by Walpole.
In Fig. XVIII, 4, A, B, C, and D are glass cylinders with plane bases standing in a
box which is painted dull black on the inside. A contains the coloured solution to
be tested (here the test solution + indicator), B contains an equal volume of water,
C contains a solution of known strength for comparison (here the standard buffer
solution + indicator), while D contains the same volume of the solution to be
tested as was originally added to A. The colour of the unknown solution is thus
compensated for.

Standard series using glass comparators. A number of devices are
manufactured which employ permanent glass standards which can be mounted
in special viewers. The BDH Lovibond Nessleriser Mark 3* is one of the simplest
of these instruments and can be used for a variety of determinations (Fig. XVIII,
5).

Fig.XVin,5

t consists essentially of a plastic case for holding two Nessler tubes vertically
e ween a reflector and a detachable rotating disc having nine apertures
con aimng a series of graded, permanent glass standards. Each disc incorporates

Manufactured by The Tintometer Ltd, Salisbury, England, and available from BDH Chemicals

Ltd, Poole, BH124NN, England.

703

XVIII, 5 QUANTITATIVE INORGANIC ANALYSIS

a series of standards designed for one particular test conducted under specified
conditions. Discs are available for many of the common determinations by
colorimetric methods and include: ammonia (with Nessler’s reagent); dissolved
oxygen (with indigo carmine); copper (with dithio-oxamide); nitrate (with
phenol-2, 4-disulphonic acid) and chlorine (with p-amino-N:N-diethyl-aniline
sulphate).

A similar device suitable for a wider range of determinations is the Lovibond
‘1000’ Comparator,* this also uses series of permanent glass colour standards
(Fig. XVIII, 6). The discs containing the nine glass colour standards fit into the
comparator, which is furnished with four compartments to receive small test-
tubes or rectangular cells, and is also provided with an opal glass screen. The disc

can revolve in the comparator, and each colour standard passes in turn in front of
an aperture through which the solution in the cell (or cells) can be observed. As
the disc revolves, the value of the colour standard visible in the aperture appears
in a special recess.

Over 300 standard tests can now be carried out using the Lovibond ‘1000’
Comparator including narrow and broad range pH determinations, and the
concentrations of many metal ions, detergents and organic compounds.

XVm, 5. DUPLICATION METHOD. This method finds its chief
application in the so-called colorimetric titration. A known volume, say 50 or 100
cm^, of the solution is placed in a Nessler cylinder (Fig. XVIII, 2) and a measured
volume of the reagent or reagents is then added. An equal volume of water (50 or
100 cm^) together with the same volume of the reagent is introduced into another
similar Nessler cylinder. For mixing the solutions both cylinders are provided
either with a glass tube on which a flattened bulb [ca. 1 cm diameter) is blown or
with a stirring-rod of which the lower end is flattened to a width of 1 cm and over a
length of several centimetres. The tubes should also be provided with black- or
brown-paper cylinders to exclude light from the sides. The colour intensities are

* Manufactured by The Tintometer Ltd, Salisbury, England, and available from BDH Chemicals
Ltd, Poole, BH 12 4NN, England.

704

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 6

compared by holding the tubes close together over a white surface, such as a sheet
. of opal glass or, better, in a Nessler tube-stand (Fig. XVIII, 3). A solution
containing a known concentration of the constituent being determined is added
to the blank solution from a burette (preferably of the micro type) until the
colours of the two solutions viewed by looking down into the tubes match. As a
rule, if the volume of the standard solution required to match the colour of the
unknown is less than about 2 per cent of the total volume, the volume change due
to the addition of the reagent may be neglected. It may, however, be allowed for
by a simple calculation, or the determination may be repeated by taking 100 — x
(or 50 - x) cm^ of water, where x is the volume of the standard solution employed
in the first titration. Several determinations should be carried out, and the
positions of the tubes should be interchanged — thus that on the right-hand side
should be put to the left of the observer and vice versa.

It must be emphasised that this method can be applied only when the colour is
independent of the mode of mixing, for in one tube a very dilute solution of the
substance to be determined is mixed with the reagent, whilst in the other tube a
comparatively concentrated solution of the substanee is mixed with a dilute
solution of the reagent. The development of colour should be practically
instantaneous and remain permanent during the time required for the
measurements; foreign substances present in the unknown should not affect the
colour. The method is, at best, only an approximate one, but has the advantage
that only the simplest apparatus is required.

XVIII, 6. BALANCING METHOD, Plunger-type colorimeters. The
plunger-type of colorimeter with two halves of the field of view illuminated by the
light passing through the unknown and standard solutions respectively was

%XVin,7

invented by J. D'uboscq of Paris in 1854. Various
improved modifications of the instrument were
subsequently developed by manufacturers of opti-
cal apparatus. Before describing the latter, reference
must be made to Hehner cylinders (Fig. XVIII, 7).
These are utilised in pairs, and are the simplest form
of apparatus employed in matching colours by the
balancing method. Each cylinder has a glass
stopcock about 2.5 cm from the bottom through
which liquid may be drawn off until the colour in the
two cylinders is the same in intensity when viewed
vertically. The cylinders are graduated at Icm^
intervals, and usually have a capacity of 100 cm^;
they should have flat, carefully ground and polished

, . bottoms of clear glass and be uniform in bore. It is

f ''■sable to place them in a box so arranged that the light is reflected from the
oottom of the latter up through the tubes.

8 T principles of a Duboscq colorimeter are illustrated in Fig. XVIII,

iiist ^ source of illumination concealed in the base of the

thr through the windows (matt white screens) in the top of the base

abs°^h a ■ ^°*^tions to be tested and through the plungers. Some of the hght is
dene^a passing through the liquids, the amount of absorption being
oflieht the concentration and the depth of the solution. The two beams

a rom the plungers are then brought to a common axis by a prism system.

705

XVin, 6 QUANTITATIVE INORGANIC ANALYSIS

On looking through the eyepiece, a wide, circular field is visible, light from one
cup illuminating one half, and light from the second cup illuminating the other
half of the field. The depths of the columns of liquids are adjusted by rotating the
milled heads on either side of the instrument, which. raises and lowers the cups,
until the two halves of the field are identical in intensity, i.e., until the dividing line
practically disappears. When this condition holds and Beer’s law is applicable,
the concentrations of the two solutions are inversely proportional to their depths,
which are normally read on the scales attached to the cup carriers. The two scales
are 60 cm long and are engraved on metal : they are divided into millimetres, and a
vernier scale enables readings to be taken to within 0.1 mm.

Use of Duboscq-type colorimeter. The colorimeter must be kept
scrupulously clean. The cups and plungers are rinsed with distilled water and

either dried with soft lens-polishing material
or rinsed with the solution to be measured.

Make sure that the readings are zero when
the plungers just touch the bottoms of the
cups. Place the standard solution in one cup,
and an equal volume of the unknown solution
in the other; do not fill the cups above the
shoulder. Set the unknown solution at a scale
reading of 10.0 mm and adjust the standard
until the fields are matched. Carry out at least
six adjustments with the cup containing the
standard solution, and calculate the mean
value.

The plungers should always remain below
the surface of the liquid. Since the eye may
become fatigued and unable to detect small
differences, it is recommended after making
adjustment to close the eyes for a moment or
to look at something else, and then see if the adjustment still appears satisfactory.
It is advisable to approach the match point both from above and below.

If /j and /j are the average readings for the cups containing the solutions of
known and unknown concentration respectively, and and C 2 are the
corresponding concentrations, then if Beer’s law holds:

cJi = C 2 I 2 or C 2 = Ci-
‘2

It will be noted that if /, = 10.0, the standard scale when multiplied by 10 will give
the percentage concentration of the sample in terms of the standard.

Owing to optical and mechanical imperfections of some makes of colorimeters,
it is sometimes found that the same reading cannot be obtained in the adjustment
for illumination when the cups are filled with the same solution and balanced. In
such a case one of the cups (say, the left one) is filled with a reference solution
(which may be a solution containing the component to be determined) of the
same colour and approximately the same intensity as the unknown and the
plunger set at some convenient point (about the middle) of the scale. Fill the other
cup with a solution having a colour corresponding to a known concentration of
the component to be determined, and adjust this cup to colour balance. Take the
reading and repeat the adjustment, say ten times, in such a way that the balancing

Fig.xvm,8

706

COLORIMETRY AND SPECTROPHOTOMETRY XVni, 7

point is approached five times from the lower and five times from the higher side.
Calculate the average reading (li). Remove the cup, rinse it thoroughly, and fill it
with the unknown solution. Repeat.the balancing exactly as for the standard
solution, and find the average of, say, 10 readings { 12 )- If ci is the concentration in
the standard solution, then the concentration of the unknown solution is given
by: : . - • . ■ .

■ _ h

, C2 — Cl ,

(This method is comparable in many respects to the method of weighing by
substitution.) If Beer’s law is not valid for the solution, it is best to arrange matters
so that the colour intensity of the standard lies close to that of the unknown.

Immediately the determination has been completed, empty the cups arid rinse
both the cups and plungers with distilled water. Leave the colorimeter in a
scrupulously clean condition;

XVin, 7 . PHOTOELECTRIC PHOTOMETER METHOD. Photoelectric
colorimeters (absorptiometers). One of the greatest advances in the design of
colorimeters has been the use of photoelectric cells to measure the intensity of the
light, thus eliminating the errors due to the personal characteristics of each
observer. Before describing the various types of photoelectric colorimeters and
spectrophotometers, a brief, account will be given of the construction and
properties of the light-sensitive devices employed. Photoemissive and barrier-
layer cells are commonly used. ■

Photoemissive cells. In the simplest form of photoemissive cell (also called

Photo-emitting
V cathode

phototube) a glass bul b is coated intern all.v-avith-a-thin„sensi tLve layer, such ns
cesium or potassjum oxide an d~sllve r oxide (i.e., one which emits electrons when
^ illuminated), a free space being left to permit the

entry of the light. This layer is the cathode. A
m metal ring inserted near’ the centre of the bulb

forms the anode, and is rinaintained-at a. high
, 0 voltage by means of a battery. The interior of the
V ^/ \ either evacuated or, less desirably,

^ ^ filled with an inert gas at low pressure (e.g.,

<J argon at about 0.2 mm). When light, penetrating

i ^ thebulb, falls on the sensitive layer, electron’s are

~p emitted, thereby causing a current to flow

through an outside circuit; this current may be
AAA/WW amplified by electronic means, and is taken as a

— ^ ^ — measure of the amount of light striking the

photosensitive surface. Otherwise expressed, .the
emission of electrons leads to a potential-drop
‘ across a high resistance in series with the cell and

Q the battery; the fall in potential may be measured

T- ¥ — by a suitable potentiometer, and is related to the

amount of light falling on the cathode. :The
action of the photoemissiye celL is shown
diagrammatically in Fig. XVIII, 9.

I ° ) The sq gsitivity of a photoemissive cell fplintr.-

tube) ma y be con'siaerablv m'creasedhv nf
Eig.XVIII,9 the so-called photomultiplier tuhe.-The latter

Amplifier

707

XVIII, .7 QUANTITATIVE INORGANIC ANALYSIS

resistance; barrier-layer cells are largely used where low cost and portability are
required. On the other hand, photoemissive cells have very high internal
resistances, and their output currents are readily amplified; they are usually
employed in the most sensitive devices measuring low intensities of illumination
and, indeed, may be ruined by high intensities of incident radiant energy. Barrier-
layer cells may exhibit fatigue effects, particularly at sudden exposure to high
levels of illumination: the current may rise to a value several per cent higher than
the apparent equilibrium value and then fall off gradually. Upon standing in the
dark, the original sensitivity is recovered. The fatigue effect can be minimised by
careful selection of the optimum level of illumination, resistance of the measuring
circuit, etc.

Light filters. Optical filters are used in colorimeters (absorptiometers) for
isolating any desired spectral region. They consist of either thin films of gelatin
containing different dyes or of coloured glass. The extensive range of Wratten

filters supplied by Kodak are of
the former type, the gelatin films
being about 0. 1 mm thick.

Other optical filters are manu-
factured by Ilford and by Corning.
The transmission curves for a
series of Ilford Standard Spectrum
Filters are depicted in Fig. XVIII,
14. The manufacturers give the
following transmission ranges for
the various filters; No. 601
Spectrum Violet, 380-470 nm. No.
602, Spectrum Blue, 440-490 nm;
No. 603, Spectrum Blue-Green,
470-520 nm. No 604, Spectrum
0 400 450 500 550 600 650 700 Green, 500-540 nm; No. 605,

Wavelength, nm Spectrum Yellow-Green, 530-570
’ nm; No. 606, Spectrum Yellow,

560-610 nm; No. 607, Spectrum Orange, 570 nm with absorption increasing
from 600 nm onwards ; No. 608, Spectrum Red, 620 nm into infrared. In addition
to the above Messrs Ilford market a series of Bright Spectrum Filters, Nos.
621-626 (No. 621 Bright Spectrum Violet to No. 626 Bright Spectrum Yellow)
which are considerably brighter (i.e., have a higher transmission) than the
standard Spectrum Filters but with a slightly wider transmission range.

Interference filters (transmission type) have somewhat narrower transmitted
bands than coloured filters and are available commercially.* Interference filters
are essentially composed of two highly reflecting but partially transmitting films
of metal (usually silver separated by a spacer film of transparent material). The
amount of separation of the metal films governs the wavelength position of the
pass band, and hence the colour of the light that the filter will transmit. This is the
result of an optical interference effect which produces a high transmission of light
when the optical separation of the metal films is effectively a half wavelength or a
multiple of a half wavelength. Light which is not transmitted is for the most part

* For example, from Bausch and Lomb Inc, 820 Linden Avenue, Rochester, New York, 14625; and
from Barr and Stroud Ltd, Caxton Street, Anniesland, Glasgow, G13 IHZ.

710

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 7

reflected. The wavelength region covered is either from 253-390 nm or from 380-
1 100 nm, peak transmission is between 25-50 per cent and the bandwidth is less
than 18 nm for the narrowband filters suitable for colorimetry.

The ideal way of selecting a filter for use with a coloured solution is to construct
first, by means of a suitable spectrophotometer, the absorption curve for the
visible spectrum. Comparison of this curve with the spectral transmission curves
of the set of filters supplied by the manufacturers enables a suitable choice to be
made. Alternatively, absorbance (or transmittance)/concentration calibration
curves may be constructed with a photoelectric colorimeter, using each of the
filters in turn. As a general rule, the best filter to use in a particular determination
is that which gives the maximum absorption or minimum transmission for- a
given concentration of the absorbing substance. Less satisfactory methods
include the use of a filter that gives the smallest transmission for a given
concentration and depth of cell, and the use of a filter whose colour is as close as
possible to the complementary colour of the solution. A table of complementary
colours is given below.

Complementary colours

Wavelength (nm)

Hue (transmitted)

Complementary hue

400-435

Violet

Yellowish-green

435-480

Blue

Yellow

480-490

Greenish-blue

Orange

490-500

Bluish-green

Red

500-560

Green

Purple

560-580

Yellowish-green

Violet

580-595

Yellow

Blue

595-610

Orange

Greenish-blue

610-750

Red

Bluish-green

Prisms. To obtain improved resolution of spectra in both the visible and
ultraviolet regions of the spectrum it is necessary to employ a better optical
system than that possible with filters. In many instruments, both manual and

automatic, this is achieved by using prisr
from incandescent tungsten or deuterium
upon the fact that the refractive index.

ns to disperse the radiation obtained
sources. The dispersion is dependent
II, of the prism material varies with
wavelength, 2, the dispersive power
being given by dn/dA. The separation
achieved between different wave-
lengths is dependent upon both the
dispersive power and the apical angle
of the prism.

In instruments in which the radi-
ation is only passed through the
prism in a single direction it is
common to use a 60° prism. In some
cases double dispersion is achieved
by reflecting the radiation back
through the prism by placing a
mirrored surface behind the prism, as

711

XVra, 7 QUANTITATIVE INORGANIC ANALYSIS

in the Littrow mounting, Fig. XVIII, 15.* Monochromatic radiation of different
wavelengths is brought to focus on the instrument slit by rotation of the prism.

Unfortunately no single material is entirely suitable for use over the full range
of 200-1000 nm, although fused silica is the favourite compromise material. Glass
prisms can be employed between 400 and 1000 nm for the visible region, but are
not transparent to ultraviolet radiation. For the region below 400 nm quartz or
fused silica prisms are required. If quartz is employed for a 60° single pass prism it
is necessary to make the prism in two halves, one half from right-handed quartz
and the other from left-handed quartz in order that polarisation effects
introduced by one will be reversed by the other.

Prisms have the advantage that, unlike the diffraction gratings described
below, they only produce a single order spectrum.

Diffraction gratings. This alternative method of dispersion uses the
principle of diffraction of radiation from a series of closely spaced lines marked on
a surface. Early diffraction gratings were made of glass through which the
radiation passed and became diffracted; these are known as transmission
gratings. To achieve the diffraction of ultraviolet radiation, however, modern
grating spectrophotometers employ metal reflection gratings with which the
radiation is reflected from the surfaces of a series of parallel grooves. These are
often known as echelette gratings.

The principle of diffraction is dependent upon the differences in path length
experienced by a wavefront incident at an angle to the individual surfaces of the
grooves of the grating. If i is the angle of incidence and r the angle of reflection the
path difference between rays from adjacent grooves is given by

d sin i - d sin r

where d is the distance between the grooves. Fig. XVIII, 16. Because of the path

712

COLORIMETRY AND SPECTROPHOTOMETRY XVUI, 8

difference that is created the new wavefronts interfere with each other except
when tl^ e path difference is an integral number of waveleng ths, i.e., when

nA = d{sini±sinr) (14)

When polychromatic radiation is incident upon the diffraction grating this
equation can usually only be satisfied for a single wavelength at a time. Rotation
of the grating to change the angle of incidence i will bring each wavelength in turn
to a position to satisfy the equation, thus serving as a method of
monochromation.

Diffraction gratings suffer from the disadvantage that they produce second-
order and "^igher^der spectra which ~can ov erlap the desir^ first-order
spectrum. T his overlarrfS~tHosl co. mmonl-v-seea-bet-weea-the-loiig wavelength
region of the~firiTorder spectru m and the shorter wavelength region of the
second-order spectrum. The'difficulty is overcome by using carefully positioned
filters in the instrument to block the undesired wavelengths.

For ultraviolet/visible spectrophotometers the gratings employed have
between 10000 and 30000 lines cm~ *. This very fine ruling means that the value
of d in equation 14 is small and produces high dispersion between wavelengths in
the first-order spectrum. Only a single grating is required to cover the region
between 200 and 900 nm. The Unicam SP1700 (Fig. XVIII, 33) is a
spectrophotometer in which monochromation is obtained by a diffraction
grating.

Instruments

XVIII, 8. PHOTOELECTRIC COLORIMETERS (ABSORPTIO-
METERS). Photoelectric colorimeters may be divided into two main classes:
one-cell and two-cell instruments. Exapiples of the former are the ‘EEL’ portable
colorimeter and the long cell absorptiometer, * the Unicam SP 300 G.P.
photoelectric colorimeter,^ and the Bausch and Lomb Spectronic 20
colorimeter,^ while an example of the latter is the Hilger Spekker absorptiometer
(type H 760).“^

The essential parts of a one-cell filter photoelectric photometer (Fig. XVIII, 17)
are a light source, a light filter, a container for the solution, a barrier-layer
photocell to receive the transmitted light, and some means for measuring the
response of the photocell. A brief description will now be given of some typical
instruments.

Corning colorimeter 252. This particular instrument, which is capable of an
accuracy within ± 1 per cent, is illustrated in Fig. XVIII, 18. It employs a series of
drop-in gelatine filters (Ilford number 601-608) to cover the wavelength range
from 400-710 nm. The transmitted radiation, from a tungsten filament lamp, is
detected by means of a phototube which provides a signal to a moving coil analog
meter. The instrument can be used with sample volumes as email as 0.8 cm^.

Manufactured by:
'Corning-EELLtd.
^Pye-Unicam Ltd.
^Bausch and Lomb Inc.
Rank Hilger Ltd.

713

XVra, 8 QUANTITATIVE INORGANIC ANALYSIS

Source

Fig. xvni, 18

714

COLORIMETRY AND SPECTROPHOTOMETRY XVni, 8

‘EEL’ absorptiometer (long cell type). This instrument, presented
diagrammatically in Fig. XVIII, 19, will accommodate cells of 2.5, 5, 10, 20, 40,
and 100 mm, optical path, corresponding to volumes of 1.5, 3, 6, 12, 24, and 60
cm^. The sliding sample-carriage has a simple lever mechanism to bring into
position the reference solution and subsequently the sample to be measured.

Control Spectrum Sensitive

Fig.xvm,19

Readings are taken on a sensitive microammeter, calibrated in percentage
transmission and absorbance. There is a colour-filter wheel which permits the
easy insertion into the light beam of any one of nine spectrum filter^ , ,

Unicam SP. 300 G.P. photoelectric colorimeter. This colorimeter (Fig.
XVIII, 20, the optical system is shown in Fig. XVIII, 21) operates from a 6-volt
battery or from the a.c. mains supply through a constant-voltage transformer.

Fig.xvni,20

There is a dual cell holder which brings either cell into the same relative position
with respect to the light path. The light is controlled by a shutter mechanism - it
passes through an absorption cell with an optical path of 10 mm, which holds the
oiution under test. The transmitted light passes through the selected filter before

715

xvnr,8

Movable

holder

Source

}

4 >

ColHmaiing I

Z3 Fiber

Solveni

COLORIMETRY AND SPECTROPHOTOMETRY XVHI, 8

‘EEL’ absorptiometer (long cell type). This instrument, presented
diagrammatically in Fig. XVIII, 19, will accommodate cells of 2.5, 5, 10, 20, 40,
and 100 mm optical path, corresponding to volumes of 1.5, 3, 6, 12, 24, and 60
cm^. The sliding sample-carriage has a simple lever mechanism to bring into
position the reference solution and subsequently the sample to be measured.

Control Spectrum Sensitive

Fig.xvm,19

Fig.XVin,20

There is a dual cell holder which brings either cell into the same relative position
With respect to the light path. The light is controlled by a shutter mechanism; it
passes through an absorption cell with an optical path of 10 mm, which holds the
solution under test. The transmitted light passes through the selected filter before

715

XVIII, 8 QUANTITATIVE INORGANIC ANALYSIS

714

COLORIMETRY AND SPECTROPHOTOMETRY XVni, 8

Control Spectrum Sensitive

Fig.xvm,19

Unicam SP. 300 G.P. photoelectric colorimeter. This colorimeter (Fig.
XVIII, 20, the optical system is shown in Fig. XVIII, 21) operates from a 6- volt
battery or from the a.c. mains supply through a constant-voltage transformer.

Fig. xvm, 20

There is a dual cell holder which brings either cell into the same relative position
with respect to the light path. The light is controlled by a shutter mechanism- it
passes through an absorption cell with an optical path of 10 mm, which holds the
solution under test. The transmitted light passes through the selected filter before

715

XVni, 8 QUANTITATIVE INORGANIC ANALYSIS

reaching the barrier-iayer photoceii, the output current of which is indicated on
the galvanometer. Ilford spectrum filters can be supplied. The galvanometer is
calibrated with a linear scale 0-100, underneath which is a logarithmic scale for
calculations of the absorbance.

Bausch and Lomb Spectronic 20 colorimeter. The instrument, shown in Fig.
XVIII, 22, consists essentially of a diffraction-grating monochromator and an
electronic detection, amplification, and measuring system. It operates from 115-

Fig, XVm, 22

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 8

volt, 60-hertz mains or from a battery. The wavelength range is from 375 to 650
nm.'and can be extended to 950 nm by the addition of a red filter and exchange; of
phototubes : the effective band- width is about 20 nm.

The optical system is presented in Fig. XVIII, 23. White light from the tungsten
lamp is focused by lens A on the entrance slit; lens B collects the light from the
entrance slit and refocuses it bn the exit slit after it has been reflected and
dispersed by the diffraction grating. To obtain various wavelengths, the grating is
rotated by means of an arm which is moved when the cam is rotated; the

wavelength scale is fastened to the same shaft as the cam. The monochromatic
light which passes through the exit slit goes on through the sample to be
measured and falls upon the phototube. Whenever the sample is removed from
the instrument, an occluder falls into the light beam so that the amplifier control
can be adjusted with no further manipulation. A light control is provided for
setting the meter to full-scale deflection with a suitable blank in the sample
compartment. Cuvettes or special small test-tubes are used as containers for the
samples.

The apparatus, although primarily designed as a colorimeter, also serves as an
inexpensive spectrophotometer. For colorimetric work, the wavelength control is
rotated until the desired wavelength in nm is indicated on the wavelength scale.
The amplifier control is adjusted to bring the meter needle to zero on the ‘Percent
Transmittance Scale’ or oo on the ‘Absorbance Scale’. The test-tube or cuvette
containing water or other solvent is then inserted in the sample holder. The light
control is then rotated until the meter reads ‘100’ or ‘O’. The unknown sample is
then inserted in place of the blank and the Percent Transmittance or Absorbance
read directly from the meter.

Two-cell instruments. In view of possible variations of the operating
current of the light source in one-cell colorimeters, two-cell circuits have been
proposed based upon the idea that fluctuations would affect the two cells equally
and thus be compensated. In addition, the null-point method of balancing the
cells against each other, as indicated by a galvanometer, is supposed largely to
eliminate errors arising from cell fatigue or temperature changes. The two
photocells should be selected on the basis of similarity in spectral response, and
should be matched as closely as possible.

Hilger Spekker absorptiometer. The actual* instrument (type H. 760) is

717

XVra, 8 QUANTITATIVE INORGANIC ANALYSIS

Fig. XVra, 24a

vvTirf optical arrangements are incorporated in Fig.

A VIII, 24b, which includes the photoelectric circuit. A 100-watt projection lamp
, mounted in a central lamphouse and run from the electric-supply mains, is the
source of hghL To the right of G the light first passes through a heat-absorbing
er H, and then through a lens I to render it parallel before it passes through a
cam-shaped diaphragm J, which controls the aperture of the beam and hence its
intensity. The light then passes through lens K to a cell containing the absorbing
iquid L, a selected colour filter M, and a lens N, which forms an image of the light
source on the surface of the photocell O. On the left of G the light passes through
a heat-absorbing filter F, symmetrical with H, an iris diaphragm E, a lens system
an (corresponding with that on the right-hand side of the instrument), a

718

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 8

selected colour filter C, aad, finally, an image of the light source G is formed on
the photocell A.

The cam-shaped disc J is connected with a large, calibrated drum, and enables
the intensity of the light falling upon the photocell O to be varied by known
amounts. Since there is an image of the filament on the cell there is no change in
the photocell area illuminated when the aperture alters ; only the quantity of light
reaching the cell is controlled by the variable aperture formed by the upper portion
of the circumference of the cam and the diaphragm limiting the area of the light
beam. The scale associated with the aperture is so calibrated that if R is the
reading corresponding to a degree of opening such that the amount of light
transmitted is 1/A of that admitted when the aperture is fully open, then R
= log A. This function, known as absorbance, was chosen because it is
approximately linear with the concentration of a solution over small ranges. This
function is normally a logarithmic one, but by giving the cam disc J a suitable
contour, an evenly divided scale on the drum has been provided; there is also a
scale of percentage transmissions.

The cell at A simultaneously receives light from G of an intensity controlled by
an iris diaphragm E; the latter is uncalibrated, and is used only for adjusting
purposes and also as a fine adjustment for the final setting. The two photocells A
and 0 are connected in opposition across a Cambridge spot galvanometer P, so
that when the photoelectric currents given by the cells are equal the galvanometer
records zero deflection. A variable resistance R is arranged to provide variable
sensitivity.

Use of Spekker photoelectric absorptiometer. This is best illustrated by
describing the procedure for making a determination. Let us suppose that it is
desired to compare the depth of colour or, more precisely stated, the amount of
light absorbed by two liquids and S 2 , the latter being the more deeply coloured,

i.e., the more absorbing.

1. Place S 2 , contained in the special cell, into the beam, and open the variable
aperture to its full extent by setting the drum at zero.

2. Adjust the iris diaphragm in front of the compensation cell until the
galvanometer shows zero deflection.

3. Substitute Sj for S 2 , when the galvanometer will be seen to be deflected.

4. Adjust the calibrated variable aperture by means of the drum until the
galvanometer returns to zero, and take the reading on the drum.

This series of operations takes less than half a minute. The cells containing the
liquids are mounted side by side, and are easily interchanged by pushing the slide
along an inch or so until it clicks into the correct position. . .

If we assume that the light intensity remains constant throughout the series of
operations, then the current given by the indicating cell at the end of operation 4
IS the same as at the end of operation 2 (since in each case it balances the output of
the compensating cell). The difference in the illumination condition in the two
cases IS that in the second case the intensity-reduction produced by closing down
the aperture is substituted for the reduction produced by the absorption of the
specimen. The ratio of the area of the partly closed aperture to that of the aperture
when fully open is thus a measure of the absorption of the liquid. Since both
photocells are affected alike by changes in the intensity of the lamp, the reading is
by changes occurring during the series of operations,
the sensitivity of the instrument in detecting small differences between the
absorption of two liquids is greatly increased by the use of an appropriate light

719

XVin, 9 QUANTITATIVE INORGANIC ANALYSIS

filter; indeed, as indicated in Fig. XVIII, 24, b, the use of a suitable light filter
should be the normal practice (for a discussion as to the choice of filter, see
previous paragraphs on Light Filters). A set of eight pairs of Ilford spectrum
filters, which have a narrow band of transmission and a fairly sharp cut-off (see
Fig. XVIII, 14), is normally supplied for use with the instrument. Sliding carriers
are provided for two filters which enable them to be interchanged quickly in the
absorpliometer or changed at will. The all-glass cells for liquids are available in
lengths of 0.25, 0.5, 1, 2, and 4 cm.

It is a well-known fact that photoelectric cells under prolonged illumination
tend to behave irregularly. This difficulty is overcome by the use of a gravity-
controlled shutter which must be held open while the readings are being taken. In
this way the steady burning of the lamp itself is ensured, and at the same time it is
impossible to expose the cells for any longer than is required for the reading to
be made.

For the routine use of the absorptiometer in colorimetric determinations, it is
necessary to prepare a calibration curve by taking readings with a number of
coloured solutions of known concentration covering the required range. This
calibration curve remains valid so long as appreciable changes do not take place
in the spectral sensitivity of the photoelectric cells or in the colour of the filter. The
changes in the photocells and in the glass of the filters are generally very gradual,
and the calibration curve need be checked only at wide intervals, say, every few
months.

The aduantages of the Spekker absorptiometer include;

(a) It runs directly from the electric mains supply; no batteries are required.

(h) Owing to the use of the two balanced photocells, the readings are largely
independent of the fluctuations of the mains supply.*

(d) The instrument readings are not affected by variations in the sensitivity of the
photocells or of the galvanometer, since a null method is employed.

(e) The galvanometer, which indicates the photoelectric current, is a robust but
sensitive instrument of the spot type, and is used as a null indicator.

(/) Readings can be taken with as little as 2.5 cm^ of liquid with the 0.5-cm cells;
if a micro-cell is employed the volume can be reduced to 0.5 cm^. The
commonly used cell ( 1 cm) has a capacity of approximately 8 cm^.

XVIII, 9. PHOTOELECTRIC SPECTROPHOTOMETERS. Spectro-
photometers, from the standpoint of analytical chemistry, are those instruments
which enable one to measure transmittance (or absorbance) at various
wavelengths. Photoelectric spectrophotometers may be regarded as refined filter
photoelectric photometers (absorptiometers) employing continuously variable
and more nearly monochromatic bands of light. The less expensive instruments,
such as the Bausch and Lomb Spectronic 20 grating colorimeter which give a
band-width of 20-30 nm, have already been described: more elaborate
spectrophotometers giving a band-width of 5-10 nm (or even less) will now be
discussed briefly.

* For prolonged experiments, the use of a constant voltage transformer in the lamp supply is
recommended.

720

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 9

Fig.XVin,25

Unlearn SP600 UV Spectrophotometer. This precision spectrophotometer
(Fig, XVIII, 25) covers the range 220-1000 nm, i.e., the ultraviolet, visible, and
near infrared. It operates on either 1 10-120 V or 200-250 V. The optical system is
shown in Fig. XVIII, 26. The main features are the tungsten and deuterium
sources, the slit system, including a slit-width indicator fitted to the slit control
knob, a Littrow monochromator with a silica prism and two high sensitivity
vacuum photocell detectors. An image of the light source is directed through the
lower half of the slits to the collimating mirror M5, and then to the silica prism

721

XVIII, 9 QUANTITATIVE INORGANIC ANALYSIS

and Littrow mirror M6. On the return path the light passes through the upper
half of the slits, and through the absorption cell to the appropriate photocell. The
spectral band-width of the instrument is less than 3 nm over most of the
wavelength range and not more than 10 nm at the extremes. The cell
compartment can accommodate four rectangular cells of light path from 1 to 40
mm. Two vacuum-type photocells are fitted, a red cell for use above ca. 620 nm
and a blue cell for shorter wavelengths. Both cells are in circuit when the
instrument is in operation, and the change from one photocell to the other is
effected by a simple control operating a plane mirror, and thus no resetting of the
dark current is necessary. The amplified output of the photocells is balanced by a
potentiometer, which is calibrated in both percentage transmission (linear) and
absorbance (logarithmic); the length of the scale is about 28 cm. The instrument is
suitable for ordinary absorption determinations in large numbers (four cells can
be accommodated) or for the plotting of absorption spectra over the range 220-
1000 nm from direct readings.

Unicam SP500 Series 2 Spectrophotometer. This is a precision
photoelectric spectrophotometer with a wide range of applications, including (a)
plotting the absorption curves of liquids throughout the visible and ultraviolet
regions, (b) determining absorption (or transmission) at any previously chosen
wavelengths, and (c) quantitative analysis of mixtures of known components by
their visible or ultraviolet absorption. Fig. XVIII, 27 depicts the actual

Fig. xvni, 27

instrument, and Fig. XVIII, 28 is a schematic diagram of the optical system. The
two light sources are a deuterium arc lamp for the ultraviolet and a tungsten-
filament lamp for the visible range. Light from the lamps is selected by the
solenoid-operated mirror Ml either automatically (at 340 nm) or manually. The

722

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 9

beam is then focused onto the entrance slit of the monochromator and is
dispersed by a 30“ rear-aluminised silica prism. Monochromatic radiation of the
required wavelength is passed through the exit slit and collimated onto the
sample by lens LI. Two vacuum photocells are employed as detectors — a red-
sensitive cell is used at wavelengths above 625 nm and a blue-sensitive cell used
for the shorter wavelengths. The cell compartment has a four-cell holder to
accommodate glass or silica cells with a light path of up to 40 mm. Standard cells
are available in glass (320-1000 nm), silica (200-1000 nm), and in ‘Siiprasil’ (186-
1000 nm), with light paths of 1, 2, 5, 10, 20, 30, and 40 mm. Stoppered cells (10 mm)
are also supplied. The power supply for the lamps is provided from an
electronically stabilised unit, fed from the mains voltage. This system is claimed to
give exceptional baseline stability.

Beckman DU ultraviolet and visible spectrophotometer. This is a precision
instrument. Two interchangeable light sources are used: a tungsten-filament
lamp and a hydrogen-discharge lamp, the former for measurements down to 320
nm and the latter for measurements in the ultraviolet region below 360 nm. It
employs a quartz prism of the Littrow type with a concave mirror of 50 cm focal
length for colliination. The slit mechanism is continuously adjustable from 0.01 to
2.0 nim: slit widths are read directly from a calibrated dial and are reproducible to
within 0.1 per cent. The wavelength range is from 210 to 1000 nm, and wavelength
scale readings are accurate to better than 0.5 nm. The band-spread of the
monochromator can, if necessary, be adjusted to less than 1 nm with the

723

XVIII, 9 QUANTITATIVE INORGANIC ANALYSIS

and Littrow mirror M6. On the return path the light passes through the upper
half of the slits, and through the absorption cell to the appropriate photocell. The
spectral band-width of the instrument is less than 3 nm over most of the
wavelength range and not more than 10 nm at the extremes. The cell
compartment can accommodate four rectangular cells of light path from 1 to 40
mm. Two vacuum-type photocells are fitted, a red cell for use above ca. 620 nm
and a blue cell for shorter wavelengths. Both cells are in circuit when the
instrument is in operation, and the change from one photocell to the other is
effected by a simple control operating a plane mirror, and thus no resetting of the
dark current is necessary. The amplified output of the photocells is balanced by a
potentiometer, which is calibrated in both percentage transmission (linear) and
absorbance (logarithmic) ; the length of the scale is about 28 cm. The instrument is
suitable for ordinary absorption determinations in large numbers (four cells can
be accommodated) or for the plotting of absorption spectra over the range 220-
1000 nm from direct readings.

Fig.XVm,27

instriunent, and Fig. XVIII, 28 is a schematic diagram of the optical system. The
two light sources are a deuterium arc lamp for the ultraviolet and a tungsten-
filament lamp for the visible range. Light from the lamps is selected by the
solenoid-operated mirror Ml either automatically (at 340 nm) or manually. The

722

COLORIMETRY AND SPECTROPHOTOMETRY XYHI, 9

it emerges from the prism. The collimating mirror focuses the spectrum in the
plane of the slits D, and light of the wavelength for which the prism is set passes
out of the monochromator through the exit (upper) slit, through the absorption
cell G to the photocell H. The photocell response is amplified and is registered on
the meter. Both battery operated (6 volt, 120 ah) and a.c. mains operated (115 or
230 volt) models are available.

Double beam spectrophotometers.

Most modern general purpose, ultraviolet/visible spectrophotometers are double
beam instruments which cover the range between about 200 and 800 nm by a
continuous automatic scanning process producing the spectrum as a pen trace on
calibrated chart paper.

In these instruments the monochromated beam of radiation, from tungsten
and deuterium lamp sources, is divided into two identical beams; one of which
passes through the reference cell and the other through the sample cell. The signal
for the absorption of the contents of the reference cell is automatically subtracted
from that from the sample cell giving a net signal corresponding to the
absorption for the components in the sample solution.

Perkin Elmer 402 Spectrophotometer. This is a highly versatile double
beam spectrophotometer which covers the range from 190 to 850 nm using a
fused silica prism for monochromation and a photomultiplier detector. The
instrument (Fig. XVIII, 31) has a continuous flow chart with a linear absorbance
scale calibrated from 0 to 1.5. By use of attenuators the absorbance scale can be
extended to 3.0 and scan times from 2 to 40 minutes are possible.

b a ^ ^’8- XVIII, 32, splitting of the beam into the reference and sample

e ms does not take place until after monochromation, and is achieved by a
° ^ mirror. Scanning is from the lower wavelength limit to the

PPer, the value for the absorption being automatically plotted as the chart
per, actuated by a servo-motor, passes tinder the pen.

automatic repeated scans and has scale
Sion controls for both the wavelength and absorbance scales. It is suitable

725

XVI1I,9 QUANTITATIVE INORGANIC ANALYSIS:

appropriate setting of the slit opening over most of the spectral range of the
instrument. Two photocells are employed: a red-sensitive phototube for use
above 600 nm and a blue-sensitive phototube for use in the range 320-625 nm
(tungsten lamp) and 210-360 nm (hydrogen lamp). The phototube current is
measured by a null method utilising a slide-wire potentiometer and an electronic
amplifier. The potentiometer is cahbrated in per cent transmission from 0 to 100
and in absorbance from 0 to 2.0; a switch is provided which increases the
sensitivity by a factor of 10, thus giving greater accuracy in reading transmission
values below 10 per cent and at the same time extending the absorbance range
from 1.0 to 3.0. Controls are provided inter alia for adjusting the dark current to
zero and the percentage transmission to 100. Four standard rectangular
absorption cells of 10 mm light path are supplied in a four-place cell holder: an

Fig.xvm,30

interchangeable cell compartment is available to accommodate either Pyrex or
silica cells with path lengths from 2 to 100 mm and with sample volumes from 0.3
to 28.5 cm^

A schematic diagram of the optical system is given in Fig. XVIII, 29 and the
instrument is shown in Fig. XVIII, 30. An image of the light source A is focused by
the condensing mirror B and the diagonal mirror C on the entrance slit at D. The
entrance slit is the lower of two slits vertically over each other. Light falling on the
collimating mirror E is rendered parallel, and is reflected towards the quartz
prism F. The back surface of the prism is aluminised, so that light refracted at the
first surface is reflected back through the prism, undergoing further refraction as

724

COLORIMETRY AND SPECTROPHOTOMETRY XYHI, 10

by means of a diffraction grating and producing a resolution of 0. 1 nm. As shown
in Fig. XVIII, 33, the instrument possesses a digital display panel for the
instantaneous reading of the absorbance values as these are measured. In this
instrument the signal is fed to a separate chart recorder to produce the complete
spectrum as a pen trace. An extra large sample compartment is built into the
instrument to enable different types of cells to be fitted easily. The complete
spectrum is produced as a continuous chart at any one of eleven different speeds.

Experimental. Colorimetric determinations

XVni,10. SOME GENERAL REMARKS UPON COLORIMETRIC
DETERMINATIONS. Visual methods have been virtually displaced for most
determinations by methods depending upon the use of photoelectric cells (filter
photometers or absorptiometers, and spectrophotometers), thus leading to
reduction of the experimental errors of colorimetric determinations. The so-
called photoelectric colorimeter is a comparatively inexpensive instrument, and
should be available in every laboratory. The use of spectrophotometers has
enabled determinations to be extended into the ultraviolet region of the
spectrum, whilst the use of chart recorders means that the analyst is not limited to
working at a single fixed wavelength.

The choice of a colorimetric procedure for the determination of a substance
will depend upon such considerations as the following:

(a) A colorimetric method will often give more accurate results at low
concentrations than the corresponding titrimetric or gravimetric procedure.
It may also be simpler to carry out.

(b) A colorimetric method may frequently be applied under conditions where no
satisfactory gravimetric or titrimetric procedure exists, e.g., for certain
biological substances.

(c) Colorimetric procedures possess advantages for the routine determination of
some of the components of a number of similar samples by virtue of the
rapidity with which they may be made: there is often no serious sacrifice of
accuracy, over the corresponding gravimetric or titrimetric procedures
provided the experimental conditions are rigidly controlled.

The criteria for a satisfactory colorimetric analysis are:

1. Specificity of the colour reaction. Very few reactions are specific for a
particular substance, but many give colours for a small group of related
substances only, i.e., are selective. By utilising such devices as the introduction of
ot er complex-forming compounds, by altering the oxidation states, and control
0 pH, close approximation to specificity may often be obtained. This subject is
discussed in detail below.

2. Proportionality between colour and concentration. For visual

CO onmeters it is important that the colour intensity should increase linearly with
ph 0^ fhe substance to be determined. This is not essential for

the°‘°^ ^^hic instruments, since a calibration curve may be constructed relating
0 instrumental reading of the colour with the concentration of the solution.

expressed, it is desirable that the system follows Beer’s law even when
P otoelectric colorimeters are used.

stable colour. The colour produced should be sufficiently

reactio ° accurate reading to be taken. This applies also to those

ns in which colours tend to reach a maximum after a time: the period of

727

XVIII, 9 QUANTITATIVE INORGANIC ANALYSIS

for a wide range of studies including reaction kinetics in addition to standard
quantitative determinations.

Unicam SP 1700 Spectrophotometer. This double beam spectrophoto-
meter can be used for the range between 190 and 850 nm, monochromation being

Fig.XVra,33

726

COLORIMETRY AND SPECTROPHOTOMETRY XYHI, 10

by means of a diffraction grating and producing a resolution of 0. 1 nm. As shown
in Fig. XVIII, 33, the instrument possesses: a digital display panel for the
instantaneous reading of the absorbance values as these are measured. In this
instrument the signal is fed to a separate chart recorder to produce the complete
spectrum as a pen trace. An extra large sample compartment is built into the
instrument to enable different types of cells to be fitted easily. The complete
spectrum is produced as a continuous chart at any one of eleven different speeds.

Experimental. Colorimetric determinations

XVffl,10. SOME GENERAL REMARKS UPON COLORIMETRIC
DETERMINATIONS. Visual methods have been virtually displaced for most
determinations by methods depending upon the use of photoelectric cells (filter
photometers or absorptiometers, and spectrophotometers), thus leading to
reduction of the experimental errors of colorimetric determinations. The so-
called photoelectric colorimeter is a comparatively inexpensive instrument, and
should be available in every laboratory. The use of spectrophotometers has
enabled determinations to be extended into the ultraviolet region of the
spectrum, whilst the use of chart recorders means that the analyst is not limited to
working at a single fixed wavelength.

The choice of a colorimetric procedure for the determination of a substance
will depend upon such considerations as the following:

(a) A colorimetric method will often give more accurate results at low
concentrations than the corresponding titrimetric or gravimetric procedure.
It may also be simpler to carry out.

(b) A colorimetric method may frequently be applied under conditions where no
satisfactory gravimetric or titrimetric procedure exists, e.g., for certain
biological substances.

(c) Colorimetric procedures possess advantages for the routine determination of
some of the components of a number of similar samples by virtue of the
rapidity with which they may be made: there is often no serious sacrifice of
accuracy over the corresponding gravimetric or titrimetric procedures
provided the experimental conditions are rigidly controlled.

The criteria for a satisfactory colorimetric analysis are:

1. Specificity of the colour reaction. Very few reactions are specific for a
particular substance, but many give colours for a small group of related
substances only, i.e., are selective. By utilising such devices as the introduction of
°|“®^‘'°™plux-forming compounds, by altering the oxidation states, and control
u pH, close approximation to specificity may often be obtained. This subject is
discussed in detail below.

2. Proportionality between colour and concentration. For visual

CO orimeters it is important that the colour intensity should increase linearly with
ch of the substance to be determined. This is not essential for

P 0 oelectric instruments, since a calibration curve may be constructed relating

e instrumental reading of the colour with the concentration of the solution,
expressed, it is desirable that the system follows Beer’s law even when
P otoelectric colorimeters are used.

stable o/ the colour. The colour produced should be sufficiently

reactL ° ^n accurate reading to be taken. This applies also to those
ns m.which colours tend to reach a maximum after a time: the period of

727

XVIII, 10 QUANTITATIVE INORGANIC ANALYSIS

maximum colour must be long enough for precise measurements to be made. In
this connection the influence of other substances and of experimental conditions
(temperature, pH, stability in air, etc.) must be known.

4. Reproducibility. The colorimetric procedure must give reproducible
results under specific experimental conditions. The reaction need not necessarily
represent a stoichiometrically quantitative chemical change.

5. Clarity of the solution. The solution must be free from precipitate if
comparison is to be made with a clear standard. Turbidity scatters as well as
absorbs the light.

6. High sensitivity. It is desirable, particularly when minute amounts of
substances are to be determined, that the colour reaction be highly sensitive. It is
also desirable that the reaction product absorb strongly in the visible rather than
in the ultraviolet; the interfering effect of other substances in the ultraviolet is
usually more pronounced.

In view of the selective character of many colorimetric reactions, it is important
to control the operational procedure so that the colour is specific for the
component being determined. This may be achieved by isolating the substance by
the ordinary methods of inorganic analysis; double precipitation is frequently
necessary to avoid errors due to occlusion and co-precipitation. Such methods of
chemical separation may be tedious and lengthy: if minute quantities are under
consideration, they may also lead to appreciable loss owing to solubility,
supersaturation, and peptisation effects. Use may be made of any of the following
processes in order to render colour reactions specific and/or to separate the
individual substances:

(a) Suppression of the action of interfering substances by the formation of
complex ions or of non-reactive complexes.

(b) Adjustments of the pH; many reactions take place within well-defined limits
of pH.

(c) Removal of the interfering substance by extraction with an organic solvent,
sometimes after suitable chemical treatment.

(d) Isolation of the substance to be determined by the formation of an organic
complex, which is then removed by extraction with an organic solvent. This
method may be combined with (a) in which an interfering ion is prevented
from forming a soluble organic complex by converting it into a complex ion
which remains in the aqueous layer.

(e) Separation by volatilisation. This method is of limited application, but gives
good results, e.g., distillation of arsenic as the trichloride in the presence of
hydrochloric acid.

(/) Electrolysis with a mercury cathode or with controlled cathode potential.

(g) Application of physical methods utilising selective absorption, chromato-
graphic separations, and ion exchange separations.

Some remarks concerning standard curves seem appropriate at this point. The
usual method of use of a filter photometer or a spectrophotometer requires the
construction of a standard curve (also termed the reference or calibration curve)
for the constituent being determined. Suitable quantities of the constituent are
taken and treated in the same way as the sample solution for the development of
colour and the measurement of the transmission (or absorbance) at the optimum
wavelength. The absorbance (log/,,//,) is plotted against the concentration: a
straight line plot is obtained if Beer’s law is obeyed. The curve may then be used
for future determinations of the constituent under the same experimental

728

COLORIMETRY AND SPECTROPHOTOMETRY XVm, 11

conditions. When the absorbance is directly proportional to the concentration,
only a few points are required to establish the line : when the relation is not linear,
a greater number of points will generally be necessary. The standard curve should
be checked at intervals. When a filter photometer is used, the characteristics of the
filter and the light source may change with time.

When plotting the standard curve it is customary to assign a transmission of
100 per cent to the blank solution (reagent solution plus water); this represents
zero concentration of the constituent. It may be mentioned that some coloured
solutions have an appreciable temperature coefficient of transmission, and the
temperature of the determination should not differ appreciably from that at
which the calibration curve was prepared.

The following procedures are arranged in alphabetical order, with cations first
(Sections 11-30), followed by the anions (Sections 31-36).*

Cations

XVin, 11. ALUMINIUM. Discussion. Among the reagents that have been
used for the colorimetric determination of aluminium are ammonium
aurintricarboxylate [aluminon) and Eriochrome Cyanine R. The latter appears to
be somewhat superior, and its use will therefore be described. At a pH of 5.9-6. 1,
zinc, nickel, manganese, and cadmium interfere negligibly, but iron and copper
must be absent. One procedure for removing interfering elements, e.g., in the
analysis of steels, is to pass the solution through a cellulose column (compare
Section VIH, 7); iron and other elements are separated by elution with a mixture
of concentrated hydrochloric acid and freshly distilled ethyl methyl ketone (8:192
v/v). The aluminium, and any nickel present, are recovered by passing dilute
hydrochloric acid (1:5 v/v) through the column.

Reagents. Eriochrome Cyanine R solution. Dissolve 0.1 g of the solid
reagent in water, dilute to 100 om^, and filter through a Whatman No. 541 filter
paper if necessary. This solution should be prepared daily.

Standard aluminium solution. Dissolve 1.319 g A.R. aluminium potassium
sulphate in water and dilute to 1 dm^ in a graduated fiask; 1 cm^ = 75 ^ug Al.

Buffer solution, concentrated. Dissolve 27.5 g ammonium acetate and 1 1.0 g
hydrated sodium acetate in 100 cm^ water: add 1.0 cm^ glacial acetic acid and
mix well.

Buffer solution, dilute. To one volume of concentrated buffer solution, add
fave volumes water and adjust the pH to 6.1 by adding acetic acid or sodium
hydroxide solution.

J^^ocedure. Transfer an aliquot of the solution (say, 20.0 cm^), containing
~ 0 /ig Al and free from interfering elements, to a 250-cm^ beaker, add 5 cm^ of
-Vo ume hydrogen peroxide and mix well. Adjust the pH of the solution to 6.0
(using either 0.2M-sodium hydroxide or 0.2M-hydrochloric acid), add 5.0 cm^

nochrome Cyanine R solution, and mix. Introduce 50 cm^ of the dilute buffer
? ^'Inte without delay to 100 cm^ in a graduated flask. Measure the
sor ance after 30 minutes with a spectrophotometer at 535 nm against a

agent blank in a 5 mm cell. For an absorptiometer, use Ilford No. 605 filter and
1 cm cells.

metals are available and in some instances a reagent will produce a
Appendix tf- ^ colorimetric determination with several metals. The table reproduced in
gives a clearer idea of the wide range of reagents available.

729

XVUI, 12 QUANTITATIVE INORGANIC ANALYSIS

Construct the calibration curve using 0, 1, 2, 3, 4, and 5 cm^ of the standard
aluminium solution.

XVm, IZ DETERMINATION OF AMMONIA. Discussion. J.Nesslerin
1856 first proposed an alkaline solution of mercury(II) iodide in potassium iodide
as a reagent for the colorimetric determination of ammonia. Various
modifications of the reagent have since been made. When Nessler’s reagent is
added to a dilute ammonium salt solution, the liberated ammonia reacts with the
reagent fairly rapidly but not instantaneously to form an orange-brown product,
which remains in colloidal solution, but flocculates on long standing. The
colorimetric comparison must be made before flocculation occurs.

The reaetion with Nessler’s reagent (an alkaline solution of potassium
tetraiodomercurate(II)) may be represented as:

2K2[HgIJ - 4 - 2NH3 = NHjHgjIj -t- 4 K 1 + NH4I

The reagent is employed for the determination of ammonia in very dilute
ammonia solutions and in water. In the presence of interfering substances, it is
best to separate the ammonia first by distillation under suitable conditions. The
method is also appUcable to the determination of nitrates and nitrites: these are
reduced in alkaline solution by Devarda’s alloy to ammonia, which is removed by
distillation. The procedure is applicable to concentrations of ammonia as low as
0.1 mgdm"'^.

Nessler’s reagent is prepared as follows. Dissolve 35 g potassium iodide in
100 cm^ water, and add 4 per cent mercury(II) chloride solution, with stirring or
shaking, until a slight red precipitate remains (about 325 cm^ are required), Then
introduce, with stirring, a solution of 120 g sodium hydroxide in 250 cm^ water,
and make up to 1 dm^ with distilled water. Add a little more mercury(II) chloride
solution until there is a permanent turbidity. Allow the mixture to stand for one
day and decant from the sediment. Keep the solution stoppered in a dark-
coloured bottle.

The following is an alternative method of preparation. Dissolve 100 g
mercury(II) iodide and 70 g potassium iodide in 100 cm^ ammonia-free water.
Add slowly, and with stirring, to a cooled solution of 160 g sodium hydroxide
pellets (or 224 g potassium hydroxide) in 700 cm^ ammonia-free water, and dilute
to 1 dm^ with ammonia-free distilled water. Allow the precipitate to settle,
preferably for a few days, before using the pale yellow supernatant liquid.

Ammonia-free water may be prepared in a conductivity-water still, or by
means of a column charged with a mixed cation and anion exchange resin (e.g.,
Permutit Bio-Deminrolit or Amberlite MB-1), or as follows. Redistil 500 cm^ of
distilled water in a Pyrex apparatus from a solution containing 1 g potassium
permanganate and 1 g anhydrous sodium carbonate; reject the first 100-cm^
portion of the distillate and then collect about 300 cm^.

Procedure. For practice in this determination, employ either a very dilute
ammonium chloride solution or ordinary distilled water which usually contains
sulficient ammonia for the exercise.

Prepare a standard ammonium chloride solution as follows. Dissolve 3.141 g
A.R. ammonium chloride, dried at 100 °C, in ammonia-free water and dilute to 1
dm^ with the same water. This stock solution is too concentrated for most
purposes. A standard solution is made by diluting 10 cm^ of this solution to 1 dm^
with ammonia-free water: 1 cm^ contains 0.01 mg of NH3.

730

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 13/14

If necessary, dilute the sample to give an ammonia concentration of 1 mg dm“ ^
(Hehner cylinders,. Fig. XVIII, 7, are useful for this dilution), and fill a 50-cm^
Nessler tube to the mark. Prepare a series of Nessler tubes containing the
following volumes of standard ammonium chloride solution diluted to 50 cm^ ;
1.0, 2.0, 3.0, 4.0, 5.0, and 6.0 cm^. The standards contain 0.01 mg NHj for each cm^
of the standard solution. Add Tcm^ of Nessler’s reagent to each tube, allow to
stand for 10 minutes, and compare the unknown with the standards in a Nessler
stand (Fig. XVIII, 3) or in a B.D.H. nesslerimeter (Fig. XVIII, 5). This will give an
approximate figure which will enable another series of standards to be prepared
and more accurate results to be obtained. •

A photoelectric colorimeter or a spectrophotometer may, of course, be used.
When 1 cm^ of the Nessler reagent is added to 50 cm^ of the sample, a blue colour
filter in the wavelength region 400-425 nm allows measurements with a 10-mm
path in the nitrogen range 20-250 /ig. Nitrogen concentrations approaching up
to 1 mg can be determined with a green colour filter or in the wavelength range
near 525 nm. The calibration curve should be prepared under exactly the same
conditions of temperature and reaction time adopted for the sample.

XVin, 13. ANTIMONY, Discussion. The procedure is based on the
formation of yellow tetraiodoantimonate(III) acid (HSbl 4 ) when antimony(III) in
sulphuric acid solution is treated with excess of potassium iodide solution.
Spectrophotometric measurements may be made at 425 nm in the visible region
or, more precisely, at 330 nm in the ultraviolet region. Appreciable amounts of
bismuth, copper, lead, nickel, tin, tungstate, and molybdate interfere.

Reagentk Potassium iodide solution. Dissolve 14.0 g A.R. potassium
iodide and 1.0 g crystallised ascorbic acid in redistilled water and dilute to 100
cm^

Standard antimony solution. Dissolve 0.2668 g A.R. antimonyl potassium
tartrate in redistilled water, add 160 cm^ concentrated sulphuric acid, and dilute
to 1 dm^ with water in a graduated flask.

Procedure. Use a solution containing 0.15-1.8 mg antimony per 100 cm^;
It should be slightly acidic with sulphuric acid (1.2-1.5M). Transfer a 10-cm^
aliquot to a 50-cm^ graduated flask, add 25 cm^ of the potassium iodide-
Mcorbic acid reagent, and dilute to the mark with 25 per cent v/v sulphuric acid.
Mix thoroughly and measure the absorbance at 425 nm or at 330 nm using a
reagent blank as reference solution.

Construct a calibration curve using appropriate volumes of the standard
antimony solution treated in the same way as for the sample solution.

Discussion. Of the numerous procedures available
or e determination of minute amounts of arsenic,* only two will be described.

arsine arsenic (0.001-0.1 mg) may be determined by volatilising the element as

mercurvmi^ and comparing the coloration formed upon discs of dry paper impregnated with
Althouah ^*iat obtained by the use of known amounts of arsenic (Gutzeit test),

laborato ®^lhod is still used in practice and suitable apparatus is available from most
Too mu T houses, it is doubtful whether the accuracy exceeds 10 per cent of the true value,
same th placed upon the rate of evolution of arsine, which is not necessarily the

spactronh ®''olution of hydrogen in the reduction apparatus. On the whole, the

complex based upon molybdenum blue or the silver diethyldithiocarbamate

> s ar superior. In all evolution methods arsenic must be in the arsenic(III) state.

731

XVIII, 14 QUANTITATIVE INORGANIC ANALYSIS

viz., the molybdenum blue method- and the silver diethyldithiocarbamate
method. Both possess great sensitivity and precision, and. are readily applied
colorimetrically or spectrophotometrically.

Molybdenum blue method. When arsenic, as arsenate, is treated with
ammonium molybdate solution and the resulting heteropoly molybdioarsenate
(arseno-molybdate) is reduced with hydrazinium sulphate or with tin(II) chloride,
a blue soluble complex ‘molybdenum blue’ is formed. The constitution is
uncertain, but it is evident that the molybdenum is present in a lower oxidation
state. The stable blue colour has a maximum absorption at about 840 nm and
shows no appreciable change in 24 hours. Various techniques for carrying out the
determination are available, but only one can be given here. Phosphate reacts in
the same manner as arsenate (and with about the same sensitivity) and must be
absent.

Both macro and micro quantities of arsenic may be isolated by distillation of
arsenic(III) chloride from hydrochloric acid solution in an all-glass apparatus in a
stream of carbon dioxide or nitrogen: a reducing agent, such as hydrazinium
sulphate, is used to reduce arsenic(V) to arsenic(IIl). The distillate may be
collected in cold water. Germanium accompanies arsenic in the distillation; if
phosphate is present in large amounts the distillate should be redistilled under the
same conditions. Another method of isolation involves volatilisation of arsenic
as arsine by the action of zinc in hydrochloric or sulphuric acid solution.
Appreciable amounts of certain reducible heavy metals, such as copper, nickel,
and cobalt, slow down the evolution of arsine, as do also large amounts of metals
that are precipitated by zinc. Copper in more than small quantities prevents
complete evolution of arsine; the error amounts to 20 per cent (for 5-10 /ig As)
with 50 mg of copper. The arsine which is evolved may be absorbed in a sodium
hydrogencarbonate solution of iodine. The absorption apparatus should be so
designed that the arsine is completely absorbed.

Reagents.* Potassium iodide solution. Dissolve 15 g of the A.R. solid in
100 cm^ water.

Tin{lT) chloride solution. Dissolve 40 g A.R. hydrated tin(II) chloride in 100
cm^ concentrated hydrochloric acid.

Zinc. Use 20-30 mesh or granulated; arsenic-free.

Iodine-potassium iodide solution. Dissolve 0.25 g iodine in a small volume
of water containing 0.4 g potassium iodide, and dilute to 100 cm^

Sodium disulphite solution. Dissolve 0.5 g of the solid reagent (Na 2 S 205 ) in
10 cm^ water. Prepare fresh daily.

Sodium hydrogencarbonate solution. Dissolve 4.2 g of the solid in 100 cm^
water.

Ammonium molybdate-liydraziniumsidphatereagent. Solution (a): dissolve
1.0 g A.R. ammonium molybdate in 10 cm^ water and add 90 cm^ of 3M-
sulphuric acid. Solution (b) : dissolve 0. 1 5 g pure hydrazinium sulphate in 100 cm^
water. Mix 10.0 cm^ each of solutions (a) and (fe) just before use.

Hydrochloric acid. This must be arsenic-free.

Standard arsenic solution. Dissolve 1.320 g A.R. arsenic(III) oxide in the
minimum volume of IM-sodium hydroxide solution, acidify with dilute

* Special pure, arsenic-free reagents are available from chemical supply houses (e.g., British Drug
Houses) and are symbolised by ‘AsT after the name of the compound ; these should be used as far as
possible in the determination and for the preparation of the above reagents.

732

COLOWMETRY AND SPECTROPHOTOMETRY XYHI, 14

hydtochlorate acid, and make lip to 1 dm^ in a graduated flask: 1 cm^ contains
1 mg of As. A solution containing 0.001 mg As per cm^ is prepared by dilution.

Procedure. The arsenic must be in the arsenic(III) state; this may be
secured by first distilling in an all-glass apparatus with concentrated
hydrochloric acid and hydrazinium sulphate, preferably in a stream, of carbon
dioxide or nitrogeii. Another method consists in reducing the arsenate (obtained
by the wet oxidation of a sample)-with potassium iodide and tin(II) chloride: the
acid concentration of the solution after dilution to 100 cm^ must not exceed 0.2-
0,5M; 1 cm^ of 50 per cent potassium iodide solution and 1 cm^ of a 40 per cent
solution of tin(II) chloride in concentrated hydrochloric acid are added, and the
mkture heated to boiling.

Transfer an aliquot portion of the arsenate solution, having a volume of 25 cm^
and containing not' more than 20 jug of arsenic, to the 50-cm^ Pyrex evolution
vessel A shown in Fig, XVIII, 34, and add sufficient concentrated hydrochloric
acid to make the total volume present in the solution 5-6 cm^, followed by 2 cm^
of the potassium iodide solution and 0.5 cm^ of the tin(II) chloride solution. Allow
to stand at room temperature for 20-30 minutes to permit the complete reduction
of the arsenate.

The tube B is loosely packed with purified glass wool soaked in lead acetate
solution (to remove hydrogen sulphide and trap acid spray), and C is a capillary
tube (4 mm external and 0.5 mm internal diameter). Place 1.0 cm^ iodine-
potassium iodide solution and 0.2 cm^ of the sodium hydrogencarbonate
solution in the narrow absorption tube D. Mix with the end of the delivery tube.

Rapidly add 2.0 g of zinc to the vessel A, immediately insert the stopper, and
allow the gases to bubble through the solution for 30 minutes. At the end of this
time the solution in D should still contain some iodine. Disconnect the delivery
tube C and leave it in the absorption tube. Add 5.0 cm^
of the ammonium molybdate-hydrazine reagent and a
drop or two of sodium disulphite solution. Heat the
resulting colourless solution in a water bath at
95-100°C, cool, transfer to a 10-cm^ graduated flask,
and make up to volume with water.

Measure the transmittance of the solution at 840 nm
or with a red filter with maximum transmission above
700 nm. Charge the reference cell with a solution
obtained by taking the iodine-iodide-hydrogen-
carbonate mixture and treating it with
molybdate-hydrazinium sulphate-disulphite as in the
actual procedure.

Construct the calibration curve by taking, say, 0,
2.5, 5.0, 7.5, and 10.0 fig As (for a final volume of 10
cm^), mixing with iodine-iodide-hydrogencarbonate

Fig.XVlIl 34 solution, adding molybdate-hydrazinium

.j,, sulphate-disulphite, and heating to 95-100.

Co procedure is recommended by the Analytical Methods

mmittee of the Society for Analytical Chemistry for the determination of small
oxidaf^^ 0* arsenic in organic matter (Ref. 7). Organic matter is destroyed by wet
dithio'°”h arsenic, after extraction with diethylammonium diethyl-

Plex' chloroform, is converted into the arsenomolybdate com-
c alter is reduced by means of hydrazinium sulphate to a molybdenum

733

XVIII, 14 QUANTITATIVE INORGANIC ANALYSIS

blue complex and determined spectrophotometrically at 840 nm and referred to a
calibration graph in the usual manner.

Silver diethyldithiocarbamate method. Arsine reacts with a solution of
silver diethyldithiocarbamate, AgS-CS-N(C 2 H 5 ) 2 , in pyridine to form a soluble
red complex, which has an absorption maximum at 540 nm. This forms the basis
of the method; the arsenic must be in the arsenic(III) state. Stibine SbHj under
similar conditions yields a red colour with maximum absorption at 510 nm, and
therefore interferes.

Reagents. See above under molybdenum blue method for zinc and tin(II)
chloride.

Lead acetate solution. Dissolve 10 g pure lead acetate in 100 cm^ distilled
water.

Silver diethyldithiocarbamate-pyridine solution. Dissolve 1.0 g pure, dry
silver diethyldithiocarbamate in 200 cm^ pure pyridine. Store in an amber bottle.

Silver diethyldithiocarbamate may be prepared as follows. To a solution of
2.25 g A.R. sodium diethyldithiocarbamate NaS-CS-N(C2H5)2,3H20 in
100 cm^ of water add, slowly and with constant stirring, a solution of 1.7 g A.R.
silver nitrate in 100 cm^ water. Both solutions should be at 8-10 "C. Collect the
lemon yellow precipitate on a sintered glass funnel, wash with about 100 cm^
cold water, and then dry in a vacuum desiccator at room temperature.

Potassium iodide solution. Dissolve 15 g A.R. potassium iodide in 100 cm^
distilled water.

All glassware used should be thoroughly cleaned with either hot concentrated
sulphuric acid or boiling concentrated nitric acid, followed by rinsing with
distilled water, and then with acetone.

Procedure. Use the apparatus shown in Fig. XVIII, 35.* The flask has a
capacity of 100 or 125 cm^, and is connected to the scrubber by means of a ground
joint; the scrubber is attached to the arsine absorber by means of a ball joint. The
arsine absorber has a calibration mark at 4.00 cm^ to ensure that the same
volume of reagent is used in each determination. Impregnate the purified glass

wool in the scrubber with lead acetate
solution; this will absorb any hydrogen
sulphide which may be subsequently
evolved. Charge the absorption tube
with 4.00 cm^ of the silver diethyl-
dithiocarbamate reagent.

Prepare a calibration curve by pipet-
ting suitable aliquots of the diluted
standard arsenic solution into a series of
clean evolution flasks fitted with stan-
dard taper necks; cover the range 0-10
fig of arsenic. To each of these diluted
aliquots, add 5 cm^ of concentrated
hydrochloric acid, 2.0 cm^ of 15 percent
potassium iodide solution and 8 drops
of tin(II) chloride solution. Swirl the
contents of the flasks, and allow them to
stand for about 15 minutes to ensure

Ball joint

-H 2 S Scrubber

Evolution flask

* Modified forms are available commercially.

734

COLORIMETRY AND SPECTROPHOTOMETRY XYIH, 15/16

complete reduction to the arsenic(III) state. Add 5.0 g pure granulated zinc to the
solution in the flask and insert the hydrogen sulphide scrubber immediately. The
evolution of arsine is 99 per cent complete in 30 minutes and virtually complete in
about 40 minutes. If necessary, dilute the liquid in the arsine absorber with pure
pyridine to the 4.00-cni^ mark and pass a gentle stream of air through the
absorber to mix the solution. Transfer the absorbing solution to a 1-cm cell and
measure the absorbance at 540 nm in a spectrophotometer. Repeat the procedure
with the remaining flasks. Plot the absorbance of each aliquot (less that of the
blank) against its arsenic content in jig.

For the actual determination of arsenic in the sample solution, follow the same
procedure as for the calibration, using two flasks, one for the sample solution and
the other for the reagent blank. From the absorbance obtained at 540 nm,
evaluate the arsenic content of the sample solution by reference to the calibration
graph previously prepared.

XVin, 15. BERYLLIUM. Discussion. Minute amounts of beryllium may
be readily determined spectrophotometrically by reaction under alkaline
conditions with 4-nitrobenzene-azo-orcinol. The reagent is yellow in a basic
medium; in the presence of beryllium the colour changes to reddish-brown. The
zone of optimum alkalinity is rather critical and narrow ; buffering with boric acid
increases the reproducibility. Aluminium, up to about 240 mg per 25 cm^, has
little influence provided an excess of 1 mole of sodium hydroxide is added for each
mole of aluminium present. Other elements which might interfere are removed
by preliminary treatment with sodium hydroxide solution, but the possible co-
precipitation of beryllium must be considered. Zinc interferes very slightly but
can be removed by precipitation as sulphide. Copper interferes seriously, even in
such small amounts as are soluble in sodium hydroxide solution. The interference
of small amounts of copper, nickel, iron and calcium can be prevented by
complexing with EDTA and triethanolamine.

Procedure. Transfer the almost neutral sample solution of beryllium
(containing 5 to 80 /rg of the element in a volume of about 10 cm^) to a 25 cm^
graduated flask, add 2.8 cm^ of 2.0M-sodium hydroxide (or more if much
aluminium is present), 5.0 cm^ of 0.64M-boric acid solution, and 6.0 cm^ of the
dye solution ( 1), dilute to the mark with distilled water, and mix well. Measure the
transmittance at 520 nm, preferably using a 2 cm cell.

Construct a calibration curve (for details, see Section XYIH, 10) using A.R.
beryllium sulphate and the experimental conditions given above: cover the range
5-80 of beryllium. Evaluate the concentration of the sample solution of
beryllium with the aid of the calibration curve.

Note. 1. Prepare the dye solution by stirring 0.025 g of 4-nitrobenzene-

^o-orcinol mechanically for several hours with O.lM-sodium hydroxide; filter
before use.

BISMUTH. Discussion. When potassium iodide solution is
3 ed to a dilute sulphuric acid solution containing a small amount of bismuth a
Js orange coloration, due to the formation of an iodobismuthate(in) ion,
s produced. The colour intensity increases with iodide concentration up to about
iodide and then remains practically constant,
sol is a sensitive one, but is subject to a number of interferences. The

u ion must be free from large amounts of lead, thallium(l) copper, tin, arsenic.

735

XVIII, 17 QUANTITATIVE INORGANIC ANALYSIS

antimony, gold, silver, platinum, palladium, and from elements in sufficient
quantity to colour the solution, e.g., nickel. Metals giving insoluble iodides must
be absent, or present in amounts not yielding a precipitate. Substances which
liberate iodine from potassium iodide interfere, for example, iron(III), the latter
should be reduced with sulphurous acid and the excess of gas boiled off, or by a 30
per cent solution of hypophosphorous acid. Chloride ion reduces the intensity of
the bismuth colour. Separation of bismuth from copper can be effected by
extraction of the bismuth as dithizonate by treatment in ammoniacal potassium
cyanide solution with a 0.1 per cent solution of dithizone in chloroform; if lead is
present, shaking of the chloroform solution of lead and bismuth dithizonates with
a buffer solution of pH 3.4 results in the lead alone passing into the aqueous
phase. The bismuth complex is soluble in a pentanol-ethyl acetate mixture, and
this fact can be utilised for the determination in the presence of coloured ions,
such as nickel, cobalt, chromium, and uranium. .

Procedure. Prepare a standard solution of bismuth by dissolving 0.100 g
pure bismuth (Johnson Matthey) in 20 cm^ concentrated sulphuric acid, and
diluting to 1 dm^ with water: 1 cm^ contains 0.1 mg Bi. Other standard solutions
may be obtained by dilution.

Treat the colourless solution (ca. 15 cm^), free from interfering substances and
about M in sulphuric acid, with 1 cm^ of 30 per cent hypophosphorous acid
solution and 1 cm^ of 10 per cent aqueous potassium iodide solution. Dilute to 25
cm^ and match the yellow colour produced against standards containing the
same concentration of sulphuric acid and hypophosphorous acid. Alternatively,
measure the absorbance at or near 460 nm.

In the extraction procedure the yellow solution is allowed to stand for 10
minutes, and then extracted with 3-cm^ portions of a 3:1 mixture by volume of
pentanol and ethyl acetate until the last extract is colourless. Make up the
combined extracts to a definite volume (10 cm^ or 25 cm^) with the organic
solvent, and determine the transmittance (460 nm) at once. Construct the
calibration curve by extracting known amounts of bismuth under the same
conditions as the sample.

Bismuth in lead. Discussion. This method is based upon the extraction of
bismuth as cupferrate by chloroform from 0. IM-acid solution: as little as 1 /ig of
bismuth can be separated from 10 g of lead.

Procedure. Dissolve a suitable weight of the sample of lead in 6M-nitric
acid : add a little 50 per cent aqueous tartaric acid to clear the solution if antimony
or tin is present. Cool, transfer to a separatory funnel, and dilute to about 25 cm^.
Add concentrated ammonia solution to the point where the slight precipitate will
no longer dissolve on shaking, then adjust the pH to 1, using nitric acid or
ammonia solution. Add 1 cm^ freshly prepared 1 per cent cupferron solution, mix,
and extract with 5 cm^ chloroform. Separate the chloroform layer, and repeat the
extraction twice with 1-cm^ portions of cupferron solution + 5 cm^ of chloroform.
Wash the combined chloroform extracts with 5 cm^ of water. Extract the bismuth
from the chloroform by shaking with two lO-cm^ portions of 1 M-sulphuric acid.
Run the sulphuric acid solution into a 25-cm^ graduated flask. Add 3 drops
saturated sulphur dioxide solution and 4 cm^ of 20 per cent aqueous potassium
iodide. Dilute to volume and measure the transmission at 460 nm.

XVni, 17. BORON. Discussion. Minute amounts of boron are usually
separated by distillation from an acid solution as methyl borate. Borosilicate

736

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 17

glass should be avoided, even for the storage of chemicals. The apparatus should
be constructed of fused, silica;* a platinum dish receiver may also be used.
Distillation may be made from a strong acid solution (sulphuric or phosphoric
acid). In the simplest apparatus methanol vapour is passed through a flask
containing the solution of the sample and is condensed and collected in an excess
of either calcium hydroxide or sodium hydroxide solution in a silica or platinum
dish. In a more efficient apparatus the methanol is made to cycle between the
sample dissolved in the acid medium and a flask containing calcium or sodium
hydroxide solution: distillation can thus be continued for several hours with only
a small amount of methanol. At the end of the distillation the contents of the
receiver in which the methyl borate was collected (which must be strongly
alkaline — a minimum of four times the theoretical amount of base) are
evaporated to dryness. The residue is used for the colorimetric determination.
Most of the reagents, e.g., quinalizarin (1,2,5,8-tetrahydroxyanthraquinone) or
l,T-dianthrimide (l,T-iminodianthraquinone) react only in concentrated
sulphuric acid solution. With the former the absorption maxima for the reagent
and its boron complex lie close together, while with the latter the maximum
absorption for the reagent is below 400 nm and for the boron complex is at 620
nm. The use of dianthrimide will accordingly be described. The colour change of
U'-dianthrimide from greenish-yellow to blue in the presence of borates in
concentrated sulphuric acid is the basis of a trustworthy method for the
determination of micro amounts of boron; the effective range of the reagent is
0.5-6 /ig and the colour is stable for several hours.

Interferences in the distillation method are fluoride and large amounts of
gelatinous silica. Fluoride interference may be overcome by the addition of
calcium chloride. Strong oxidising agents, such as chromate and nitrate, interfere,
since they destroy the reagent. Boron in natural waters can be determined
without separation; the residue obtained after evaporation to dryness with a little
calcium hydroxide solution may be used directly in the colour formation. In the
analysis of steel by dissolution in sulphuric acid no oxidising compounds are
formed which can interfere with the reaction.

Reagents. Dianthrimide reagent solution. Dissolve 150 mg of 1,T-
dianthrimide in 1000 cm^ concentrated sulphuric acid {ca. 96 per cent w/w). Keep
in the dark and protected from moisture.

Standard boron solution. Dissolve 0.7621 g A.R. boric acid in water and
dilute to 1 dra^. Take 50 cm^ of this solution and dilute to 1000 cm^; the resulting
solution contains 6.667 jug B per cm^.

Dilute sulphuric acid. Prepare a 1 : 3 v/v solution.

Procedure (boron in steel). Dissolve about 3 g of the steel (B content 0.02
per cent), accurately weighed, in 40 cm^ dilute sulphuric acid in a 150-cm^ Vicor
or silica flask fitted with a reflux condenser. Heat until dissolved. Filter through a
quantitative filter paper into a 100-cm^ graduated flask. Wash with hot water,
00 to room temperature, and dilute to the mark with water. This flask (A)
0 ams the acid-soluble boron.

^ platinum crucible, fuse with 2.0 g of A.R. anhydrous sodium
sulnh”^^^’ ®oolt in 40 cm^ of dilute sulphuric acid, and add 1 cm^ of

P urous acid solution (about 6 per cent) to reduce any iron(III) salt, etc.,

g Vycor glass, containing 96 per cent of silica, is usually suitable.

737

XVin, 18 QUANTITATIVE INORGANIC ANALYSIS

formed in the fusion, and filter if necessary. Transfer the solution to a 100-cm^
graduated flask, dilute to the mark, and mix. This flask (B) contains the acid-
insoluble boron.

Transfer 3.0 cm^ of solutions A and B to two dry, glass-stoppered conical flasks
(Vycor or silica). Add 25 cm^ of dianthrimide reagent solution to each with
shaking, and insert the glass stoppers loosely. For the blank use 3.0 cm^ of
solutions A and B in two similar 50-cm^ conical flasks and add 25 cm^
concentrated sulphuric acid (98 per cent w/w). Heat all four flasks in a boiling
water bath for 60 minutes. Cool to room temperature and measure the
absorbance of each of the solutions at 620 nm against pure concentrated
sulphuric acid in 1-cm or 2-cm cells. Correct for the blanks.

To construct the calibration curve, run 5-50 cm^ of the standard boron
solution by means of a burette into 100-cm^ graduated flasks, add 30 cm^ of dilute
sulphuric acid, and make up to volume. These solutions contain 1-10 //g of B per
3 cm^. Use 3 cm^ of each solution and of a boron-free comparison solution and
proceed as above. Plot a calibration curve relating absorbance and boron content.

Calculate the total boron content of the steel (i.e., acid-soluble plus acid-
insoluble boron).

An alternative method for the determination of boron is given under Section
XVI, 8.

XVm, 18. CHROMIUM. Discussion. Small amounts of chromium (up to
0.5 per cent) may be determined colorimetrically in alkaline solution as
chromate; uranium and cerium interfere, but vanadium has little influence. The
transmittance of the solution is measured at 365-370 nm or with the aid of a filter
having maximum transmission in the violet portion of the spectrum. The
standard solution used for the preparation of the reference curve should have the
same alkalinity as the sample solution, and should preferably have the same
concentration of foreign salts. Standards may be prepared from A.R. potassium
chromate.

A more sensitive method is to employ 1,5-diphenylcarbazide
COlNH-NHCgHjl^; in acid solution (ca. 0.2M) chromates give a soluble violet
compound with this reagent.

Molybdenum(VI), vanadium(V), mercury, and iron interfere; permanganates,
if present, may be removed by boiling with a little ethanol. If the ratio of
vanadium to chromium does not exceed 10: 1, nearly correct results may be
obtained by allowing the solution to stand for 10-15 minutes after the addition of
the reagent, since the vanadium-diphenylcarbazide colour fades fairly rapidly.
Vanadate can be separated from chromate by adding oxine to the solution and
extracting at a pH of about 4 with chloroform; chromate remains in the aqueous
solution. Vanadium as well as iron can be precipitated in acid solution with
cupferron and thus separated from chromium(III).

Procedure. Prepare a 0.25 per cent solution of diphenylcarbazide in 50 per
cent acetone as required. The test solution may contain from 0.2 to 0.5 part per
million of chromate. To about 1 5 cm^ of this solution add sufficient 3A/-sulphuric
acid to make the concentration about O.IM when subsequently diluted to 25 cm^
add 1 cm^ of the diphenylcarbazide reagent and make up to 25 cm^ with water.
Match the colour produced against standards prepared from O.OOlAi-potassium
dichromate solution. A green filter having the transmission maximum at about
540 nm may be used.

738

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 19

Chromium in steel. Discussion. ■ The chromium in the steel is oxidised by
perchloric acid to the dichromate ion; the colour of which is intensified by
iron(III) perchlorate which is itself colourless. The coloured solution is compared
with a blank in which the dichromate is reduced, with ammonium iron(II)
sulphate. The method is not subject to interference by iron or by moderate

amounts of alloying elements usually present in steel.-

Procedure. Place a 1.000 g sample of the steel (Cr content < 0. 1 per cent) in
a 100 cm^ beaker and dissolve it in 10 cm^ of dilute nitric acid (1 : 1) and 20 cm^ of
A.R. perchloric acid (sp. gr. 1.70; 70-72 per cent). [If the Cr content is 0.1-1 per
cent, dissolve a 0.5000 g sample in 10 cm^ of dilute nitric acid (1 : 1) and 15 cm^ of
perchloric acid (sp. gr. 1.70).] Evaporate to dense fumes of perchloric acid and boil
gently for 5 minutes to oxidise the chromium. Cool the beaker and contents
rapidly, dissolve soluble salts by adding 20 cm^ of water, transfer the solution
quantitatively to a glass-stoppered 50 cm^ graduated flask, and dilute to the
mark. Remove an aliquot portion to an absorption cell, reduce it with a little (ca.
20 mg) A.R. ammonium iron(II) sulphate, and adjust the colorimeter or
spectrophotometer so that the reading with this solution is zero; a violet filter
having a maximum transmission between 410 and 480 nm may be used in the
colorimeter. Discard the solution in the absorption cell, and refill it with an equal
volume of the oxidised solution: the reading is a measure of the colour due to the
dichromate.

Standardisation may be carried out by the use of solutions prepared from a
chromium-free standard steel and standard potassium dichromate solution.
After dissolution of the standard steel, the solution is boiled with perchloric acid,
potassium dichromate added and the resulting solution is diluted to volume, and
measurements are carried out as above; The chromium content of any unknown
steel may then be deduced from the colorimeter reading.

XVni, 19. COBALT. Discussion. An excellent method for the colorimetric
determination of minute amounts of cobalt is based upon the soluble red complex
salt formed when cobalt ions react with an aqueous solution of nitroso-R-salt
(sodium l-nitroso-2-hydroxynaphthalene-3,6-disulphonate). Three moles of the
reagent combine with 1 mole of cobalt.

The cobalt complex is usually formed in a hot acetate-acetic acid medium.
After the formation of the cobalt colour, hydrochloric acid or nitric acid is added
to decompose the complexes of most of the other heavy metals present. Iron,
copper, cerium(IV), chromium(III and VI), nickel, vanadyl vanadium, and copper
interfere when present in appreciable quantities. Excess of the reagent minimises
the interference of iron(II), iron(III) can be removed by diethyl ether extraction
from a hydrochloric acid solution. Most of the interferences can be eliminated by
treatment with potassium bromate, followed by the addition of an alkali fluoride.
Cobalt may also be isolated by dithizone extraction from a basic medium after
copper has been removed (if necessary) from acidic solution. An alumina column
may also be used to adsorb the cobalt nitroso-R-chelate anion in the presence of
perchloric acid, the other elements are eluted with warm IM-nitric acid, and
finally the cobalt complex with IM-sulphuric acid, and the absorbance measured
at 500 nm.

Procedure. The test solution should contain between 0.001 and 0.02 mg of
cobalt. Evaporate almost to dryness, add 1 cm^ of concentrated nitric acid, and
continue the evaporation just to dryness to oxidise any iron(II) which may be

739

XVIII, 20 QUANTITATIVE INORGANIC ANALYSIS

present. Dissolve the residue in 10 cm^ of water containing 0.5 cm^ each of 1 : 1
hydrochloric acid and 1:10 nitric acid. Boil for a few minutes to dissolve any solid
material. Add 2.0 cm^ of a 0.2 per cent aqueous solution of nitroso-R-salt and also
2.0 g of hydrated sodium acetate. The pH of the solution should be close to 5.5;
check with bromocresol green indicator or with a pH meter. Boil for 1 minute,
add 1.0 cm^ of concentrated hydrochloric acid, and boil again for 1 minute. Cool
to room temperature, dilute to 25 cm^ in a graduated flask; and compare the
colour with standards or use a spectrophotometer. Determine the absorbance at
425 nm against a reagent blank. In the presence of 2 mg or more of iron it is best to
make the measurement at 500 nm to reduce the error resulting from the
absorption of light by the yellow solution.

Standard solutions may be conveniently prepared with spectroscopically pure
cobalt (Johnson, Matthey).

Cobalt in steel. Discussion. An alternative, but less sensitive, method
utilises 2-nitroso-l-naphthol, and this can be used for the determination of cobalt
in steel. The pink cobalt(III) complex is formed in a citrate medium at pH 2.5-5.
Citrate serves as a buffer, prevents the precipitation of metallic hydroxides, and
complexes iron(III) so that it does not form an extractable nitrosonaphtholate
complex. The cobalt complex forms slowly (ca. 30 minutes) and is extracted with
chloroform.

Procedure. Prepare the 2-nitroso-l-naphthol reagent by dissolving 1.0 g of
the solid in 100 cm^ of glacial acetic acid. Add 1 g of activated carbon: shake the
solution before use and filter off" the required volume.

Dissolve a known weight (ca. 0.5 g) of the steel by any suitable procedure. Treat
the acidic sample solution ( < 200 /ig Co), containing iron in the iron(II) state, with
10-15 cm^ of 40 per cent (w/v) sodium citrate solution, dilute to 50-75 cm^ and
adjust the pH to 3-4 (indicator paper) with 2M-hydrochloric acid or sodium
hydroxide. Cool to room temperature, add 10 cm^ of 3 per cent (10-volume)
hydrogen peroxide and, after 3 minutes, 2 cm^ of the reagent solution. Allow to
stand for at least 30 minutes at room temperature. Extract the solution in a
separatory funnel by shaking vigorously for 1 minute with 25 cm^ of chloroform:
repeat the extraction twice with lO-cm^ portions of chloroform. Dilute the
combined extracts to 50 cm^ with chloroform and transfer to a clean separatory
funnel. Add 20 cm^ of 2AJ-hydrochloric acid, shake for 1 minute, run the
chloroform layer into another separatory funnel, and shake for 1 minute with 20
cm^ of 2M-sodium hydroxide. Determine the absorbance of the clear chloroform
phase in a 1-cm cell at 530 nm.

For the preparation of standard cobalt solutions, use A.R. cobalt(II) chloride
or spectroscopically pure cobalt dissolved in hydrochloric acid; subject solutions
containing 0, 5, 10, 25, 50, 100, 150, and 200 fig of Co to the whole procedure.

XVIII, 20. COPPER. Discussion. Small quantities of copper may be
determined by the diethyldithiocarbamate method (Section VI, 9) or by the ‘neo-
cuproin’ method (Section VI, 10), an extraction being necessary in both cases.
In another somewhat simpler procedure, the copper is complexed with
biscyclohexanone oxalyldihydrazone and the resulting blue colour is measured by
a suitable spectrophotometer within the range 570-600 nm. The solution
measured should contain not more than 100 fig of copper.

Reagents. Bicyclohexanone oxalyldihydrazone solution (copper reagent).
Dissolve 0:1 g of the solid reagent in 10 cm^ ethanol (or industrial methylated

740

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 21

spirit) and 10 cm^ hot water, and dilute to 200 cm^. Filter, if necessary.

Synthetic standard solution (for analysis of steel). Dissolve an appropriate
weight of pure, iron (Johnson Matthey) in a mixture of equal volumes of
concentrated hydrochloric acid and concentrated nitric acid; with this solution
as base, add a suitable amount of copper nitrate solution containing 0.01 g copper
per dm^.

Procedure (copper in steel). Weigh out accurately a 0.1 -g sample of the
steel* into a 150-cm^ conical beaker, add 5 cm^ concentrated hydrochloric acid
and 5 cm^ concentrated nitric acid, and warm gently. In the presence of
interfering amounts of chromium, add 5 cm^ perchloric acid, sp. gr. 1.70, and
evaporate until strong fuming occurs. When the sample has dissolved or after the
fuming with perchloric acid, cool, add 50 cm^ cold distilled water, followed by 10
cm^ acid solution (DHCl/HNOj). Carefully add 10 cm^ concentrated
ammonia solution, sp. gr. 0.88, cool to room temperature, and dilute to 100 cm^
in a graduated flask. Return the solution to the original beaker and transfer a
10-cm^ aliquot to a 100-cm^ graduated flask. Add 20 cm^ of the copper reagent,
dilute to 100 cm^ with distilled water, and transfer to a 100-cm^ dry beaker. Allow
to stand for 10-15 minutes, and then measure the absorbance with a spectro-
photometer.

Construct a calibration curve using the synthetic standard solution; add the
standard copper solution immediately before the reagent.

XVni, 21. IRON. Three procedures will be described — the thiocyanate,- the
1,10-phenanthroline and the thioglycollic acid methods.

A. Thiocyanate method. Discussion. Iron(III) reacts with thiocyanate to
give a series of intensely red-coloured compounds, which remain in true solution;
iron(II) does not react. Depending upon the thiocyanate concentration, a series of
complexes can be obtained; these complexes are red and can be formulated as
[Fe(SCN)„]^“", where n = 1, ... 6. At low thiocyanate concentration the
predominant coloured species is [Fe(SCN)]^’*' (Fe^"^ -)-SGN”->- [Fe(SCN)]^'''), at
O.lM-thiocyanate concentration it is largely [Fe(SCN) 2 ]‘*', and at very high
thiocyanate concentration it is [FefSCN)^]^ “ . In the colorimetric determination a
large excess of thiocyanate should be used, since this increases the intensity and
also the stability of the colour. Strong acids (hydrochloric or nitric acid —
concentration 0.05-0.5M) should be present to suppress the hydrolysis;

Fe3 + -FJHjO ^Fe(OH)3 -t-3H+

Sulphuric acid is not recommended, because sulphate ions have a certain
tendency to form complexes with iron(III) ions. Silver, copper, nickel, cobalt,
titanium, uranium, molybdenum, mercury (>1 g dm “^), zinc, cadmium, and
bismuth interfere. Mercury(I) and tin(II) salts, if present, should be converted into
the mercury(II) and tin(IV) salts, otherwise the colour is destroyed. Phosphates,
arsenates, fluorides, oxalates, and tartrates interfere, since they form fairly stable
complexes with iron(III) ions; the influence of phosphates and arsenates is
reduced by the presence of a comparatively high concentration of acid.

When large quantities of interfering substances are present, it is usually best to
proceed in either of the following ways; (i) remove the iron by precipitation with a

* Chemical Standards may be used for practice in this determination- B C S

iNo. 402 (0.23 per cent Cu); No. 410 (0.47 per cent Cu); No. 406 (0.32 per cent Cu).

741

XVIII, 21 QUANTITATIVE INORGANIC ANALYSIS

slight excess of ammonia solution, and dissolve the precipitate in dilute
hydrochloric acid; (ii) extract the ‘iron(III) thiocyanate’ three times either with
pure diethyl ether or, better, with a mixture of pentanol and pure diethyl ether
(5 : 2) and employ the organic layer for the colour comparison.

Reagents. Prepare the following solutions:

Standard solution of iron(III). Use method (a), (b) or (c). (a) Dissolve
0.7022 g A.R. ammonium iron(II) sulphate in 100 cm^ water, add 5 cm^ of 1 :5
sulphuric acid, and run in cautiously a dilute solution of potassium
permanganate (2 g dm “ until a slight pink coloration remains after stirring well.
Dilute to 1 dm^ and mix thoroughly. 1 cm^ s 0.1 mg of Fe. (b) Dissolve 0.864 g
A.R. ammonium iron(III) sulphate in water, add 10 cm^ concentrated
hydrochloric acid and dilute to 1 dm^. 1 cm^ s 0.1 mg of Fe. (c) Dissolve 0.1000 g
of electrolytic iron or pure iron wire in 50 cm^ of 1:3 nitric acid, boil to expel
oxides of nitrogen, and dilute to 1 dm^ with de-ionised water.

Potassium thiocyanate solution. Dissolve 20 g A.R. potassium
thiocyanate in 100 cm^ water: the solution is ca. 2M.

Procedure. Dissolve a weighed portion of the substance in which the
amount of iron is to be determined in a suitable acid, and evaporate nearly to
dryness to expel excess of acid. Dilute slightly with water, oxidise the iron to the
iron(III) state with dilute potassium permanganate solution or with a little
bromine water, and make up the liquid to 500 cm^ or other suitable volume. Take
40 cm^ of this solution and place in a 50-cm^ graduated flask, add 5 cm^ of the
thiocyanate solution and 3 cm^ of 4M-nitric acid. Add de-ionised water to dilute
to the mark. Prepare a blank using the same quantities of reagents. Measure the
absorbance of the sample solution in a spectrophotometer at 480 nm. Determine
the concentration of this solution by comparison with values on a reference curve
obtained in the same way from different concentrations of the standard iron
solution.

B. 1,10-Phenanthroline method. Discussion. Ironfll) reacts with 1,10-
phenanthroline to form an orange-red complex [(Ci 2 H 8 N 2 ) 3 Fe]^'''. The colour
intensity is independent of the acidity in the pH range 2-9, and is stable for long
periods. Iron(III) may be reduced with hydroxylammonium chloride or with
hydroquinone. Silver, bismuth, copper, nickel, and cobalt interfere seriously, as
do perchlorate, cyanide, molybdate, and tungstate. The iron-phenanthroline
complex (as the perchlorate) may be extracted with nitrobenzene and measured
at 515 nm against a reagent blank.

Both iron(II) and iron(III) can be determined spectrophotometrically: the
reddish-orange iron(II) complex absorbs at 515 nm, and both the iron(II) and the
yellow iron(III) complex have identical absorption at 396 nm, the amount being
additive. The solution, slightly acid with sulphuric acid, is treated with 1,10-
phenanthroline, and buffered with potassium hydrogenphthalate at a pH of 3.9:
the reading at 396 nm gives the total iron and that at 515 nm the iron(II).

Reagents. Prepare the following solutions: 1,10-phenanthroline, 0.25 per
cent solution of the monohydrate in water; sodium acetate, 0.2M and 2M.
Hydroxylammonium chloride, 10 per cent aqueous solution, or hydroquinone, 1
per cent solution in an acetic acid buffer of pH, ca. 4.5 (mix 65 cm^ ofO.lM-acetic
acid and 35 cm^ of 0. 1 Af-sodium acetate solution), is prepared when required.

Procedure. Take an aliquot portion of the unknown slightly acid solution
containing 0.1-0.5 mg iron and transfer it to a 50-cm^ graduated flask.
Determine, by the use of a similar aliquot portion containing a few drops of

742

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 21

bromophenol blue, the volume of sodium acetate solution required to bring the
pH to 3.5 ± 1.0. Add the same volume of acetate solution to the original aliquot
part and then 4 cm^ each of thehydroquinone and 1,10-phenanthroline solutions.
Make up to the mark with distilled water, mix well, and allow to stand for 1 hour
to complete the reduction of the iron.. Compare the mtensity of the colour
produced with standards, similarly prepared, in’ any convenient way. If a
colorimeter is employed, use a filter showing maximum transmission at 480-520
nm, for a spectrophotometer, use a wavelength of 515.nm.

The iron may also be reduced with hydroxylanimoniurri chloride. Add 5 cm^ of
the 10 per cent hydroxylammonium solution, adjust the pH of the slightly acid
solution to 3-6 with sodium acetate, then add 4 cm^ of the 1,10-phenanthroline
solution, dilute to 50 cm^, mix, and measure the absorbance after 5-10 minutes.

C. Thioglycollic acid method. Discussion. The use of thiogly collie acid
(mercaptoacetic acid) for the determination of iron (Refs. 8 and 9) is of importance
because it is relatively free from interferences giving a red-purple colour with
Fe^ * which can be measured at 535 nm. Precipitation of AH and Cr^ ions is
prevented by the addition of ammonium citrate. The reaction with the Fe^ ions
is represented as:

Fe^^ -b2HSCH2COO- -b30H- = Fe(OH){SCH2COb)2=^- -b2H20

Reagents. Prepare the following solutions : '

Thioglycollic acid solution. Dissolve 10 cm^ of analytical grade
thioglycollic acid in water and dilute to 100 cm^.

Ammonium hydroxide solution. Take 25 cm^ of 0.880 ammonium
hydroxide and dilute to 100 cm^.

Ammonium citrate solution. Dissolve 20 g analytical grade ammonium
citrate (iron free) in water and dilute to 100 cm^.

Standard iron solution. Dissolve O.lOO g pure iron wire in the miniirium
quantity of boiling 2M-nitric acid and dilute the resulting solution to 1 dm^,
giving a solution containing 100 p.p.m. of iron.

Procedure. Take 10 cm^ of the standard solution and dilute to 100 cm^ in a
graduated flask, to give a 10 p.p.m. solution. From this solution prepare a series of
standards containing 1, 2, 3, 4, and 5 p.p.m. by taking 5, 10, 15, 20 and 25 cm^
aliquots and placing in 50 cm^ graduated flasks. Add 5 cm^ of the thioglycolhc
acid solution, 2 cm^ of the 20 per cent ammonium citrate solution and 5 cm^ of
the dilute ammonium hydroxide solution to each flask. Dilute with water and mix
to make up to the 50 cm^ mark. Using a spectrophotometer, measure the
absorbances of the solutions at 535 nm by comparison with a blank prepared in
the same manner. Plot a concentration/absorbance line from the values obtained.

Take sufficient of the sample under examination to contain approximately 0. 1 g
iron and dissolve in the minimum amount of dilute nitric acid. If an organic
material is being studied or it does not dissolve in dilute nitric acid, heat the
sample with a small 'quantity of concentrated nitric or sulphuric acid and
evaporate nearly to dryness. Take the residue into solution with a little dilute
nitric acid. Dilute the resulting solution to 1 dm^. Take 1 cm^ of this solution and
place in a 50-cm^ graduated flask and add 5 cm^ thioglycollic acid solution, 2 cm^
animonium citrate solution and 5 cm^ ammonium hydroxide. Dilute with water
and make up to the 50 cm^ mark as with the standard solutions. •

Measure the absorbance of this solution at 535 nm and compare , the value
obtained with the standard line previously plotted. From the. value for the

743

XVra, 22/23 QUANTITATIVE INORGANIC ANALYSIS

concentration of iron in solution calculate the original amount of iron in the
starting material.

XVin, 22. LEAD. Discussion. For the determination of small amounts of
lead (0.005-0.25 mg) advantage is taken of the fact that when a sulphide is added
to a solution containing lead ions a brown colour, due to the formation of
colloidal lead sulphide, is produced.

Interference is caused by the presence of; (a) neutral salts, such as ammonium
chloride and particularly tartrates and citrates, and (b) other elements, such as
copper, bismuth, iron, and aluminium. Errors due to (a) may be allowed for by
ensuring that the standards for comparison contain amounts of salts
approximately equal to those in the solution under test, while those produced by
(b) are eliminated by the usual analytical procedure or by the use of
diphenylthiocarbazone (see below). The disturbing effect due to copper and iron if
present in small amount may be overcome by the addition of a few drops of a 10
per cent aqueous solution of potassium cyanide; aluminium may be retained in
ammoniacal solution by the addition of ammonium citrate solution, a
corresponding amount of the latter being added to the standards.

Reagent Prepare a standard lead solution by dissolving either 0. 1 83 g A.R.
lead acetate crystals or 0.160 g A.R. lead nitrate in 100 cm^ water; 10 cm^ of this
are diluted to 100 cm^ for a working solution: the latter contains 0.1 mg of Pb
per cm^.

Procedure. As an illustration of the simple lead sulphide procedure, the
determination of lead in commercial tartaric acid will be described. Dissolve 10 gof
the tartaric acid in about 40 cm^ of hot water, add 5 to 6 drops of 10 per cent
potassium cyanide solution (danger!) then 25 cm^ of 1:2 ammonia solution
(which should make the solution ammoniacal), and finally, add 0.5 cm^ of 10 per
cent (w/v) sodium sulphide solution. Make up to 100 cm^. Prepare a blank
solution using 10 g lead-free tartaric acid and treat it in exactly the same way.
Measure the absorption of the two solutions with a spectrophotometer at a
wavelength of 430 nm.

Construct a calibration curve by adding 0, 0.25, 0.50, 0.75, and 1.00 cm^ of the
standard lead solution to the solution prepared as above from 10 g of lead-free
tartaric acid.

Minute amounts of lead may also be determined by the dithizone method (see
Section VI, 13). The procedure may be rendered fairly specific by first removing
the lead with sodium diethyldithiocarbamate (compare Section VI, 9) at pH 7,
and extracting the lead diethyldithiocarbamate with a mixture of equal volumes
of pentanol and ‘sulphur-free’ toluene. The organic layer is treated with dilute
hydrochloric acid, whereupon the lead complex passes into the aqueous layer.
The latter is mixed with ammoniacal dithizone solution and the lead dithizonate
extracted with carbon tetrachloride; the absorbance is measured at 515 nm. The
pH of 7 is attained by the use of a buffer solution composed of 25 g sodium citrate,
4 g sodium hydrogencarbonate, and 100 cm^ water; the sodium diethyl-
dithiocarbamate is not stable in the presence of this buffer, and so the appro-
priate amount of the solid reagent is added as required,

XVin, 23. MAGNESIUM. Two methods are commonly used for the
determination of magnesium. Titan Yellow may be used to obtain a coloured
colloidal suspension, or Solochrome Black to give a red soluble complex. In most
cases the second of these is to be preferred.

744

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 23

A. Titan Yellow method. Discussion. .When magnesium, hydroxide is
precipitated with sod iu m hydr oxide solution in the presence of the organic
dyestuff Titan Yeiiafftth gsodium saj t o f dehvdrothio-P-toluidm e sulp honic acid ;
Colour Index No. 195405^a red lal^ is formed (pH > 12). FairlTstable suspen-
sions of the lake can be obtained if the magnesium concentration is below 4-5
parts per million, and fading of the colour is prevented by the presence of
hydroxyiammonium chloride. The reagent alone gives a yellow-brown colour in
sodium hydroxide solution.

Many metals interfere, particularly those which give insoluble hydroxides in
alkali hydroxide solution (e.g. cadmium, nickel, and cobalt) or those of which
the hydroxides are soluble in an excess of sodium hydroxide, such, as aluminium,
zinc, and tin. Appreciable amounts of phosphate destroy the colour, calcium
intensifies the colour of the magnesium lake; errors due to calcium can be reduced
by adding the same amount of a calcium salt to the sample and standard
solutions.

Procedure. Remove all interfering elements, for example, iron, aluminium,
and phosphate by double precipitation with aqueous ammonia solution; also
calcium (if present in quantity) with ammonium oxalate solution, and other
metals by appropriate methods. Evaporate the filtrate to dryness to expel
ammonium salts, moisten the residue with a few drops of dilute hydrochloric acid,
and dilute to volume in a graduated flask of suitable size.

Transfer into a 50-cm^ graduated flask a volume of the sample solution (say, 25
cm^) containing 5 parts per million or less of magnesium. Almost neutralise any
acid present with dilute sodium hydroxide solution. Dilute to about 35 cm^, and
add I.O cm^ of 5 per cent aqueous hydroxyiammonium chloride solution and 1.00
cm^ of a 0.15 per cent aqueous solution of Titan Yellow (special quality for
analytical use). Then add 5.0 cm^ IM-sodium hydroxide while swirling the flask.
Dilute to the mark with water, mix, and measure the intensity of the colour of the
colloidal suspension by suitable means soon after the lake has been formed. Use a
green filter (having a maximum transmission at 535 nm) with a filter photometen

To construct the calibration curve, use standard magnesium solutions
(prepared from pure magnesium and dilute hydrochloric acid) containing 0, 1, 2,
3, 4, and 5 fig of magnesium, and treat each solution in the same manner as the
sample solution.

B. Solochrome Black method. Discussion. The difficulties inherent in the
colloidal systems involved in ‘lake’ methods may be avoided by the use of organic
reagents which form soluble coloured complexes with magnesium in basic
solution. One such reagent is Solochrome Black, which forms red soluble
complexes with magnesium. The colour is not stable; calcium, copper,
manganese, iron, aluminium, cobalt, nickel, etc., interfere. By buffering at pH
10-1, a single complex is formed by one magnesium ion and two molecules of the

Calcium may be separated from magnesium by precipitation as sulphate in
the presence of a large excess of methanol.

Reagents. Buffer solution (pH = 10.1). This consists ofa 0.75 per cent w/v
solution of A.R. ammonium chloride in dilute ammonia solution, prepared by

mixing 5 volumes of concentrated ammonia solution (sp. gr. 0.88) and 95 volumes
of water.

Solochrome Black solution. Prepare a 0.1 per cent solution in methanol;
warm to speed solution and filter.

745

XVIII, 24 QUANTITATIVE INORGANIC ANALYSIS

Procedure. Transfer the neutral sample solution (< 100 fig Mg), free from
calcium and other metals, to a 100-cm^ graduated flask with calibrated neck. Add
25 cm^ of the buffer solution, dilute to just below the 90-cm^ graduation mark,
and shake. Add 10.0 cm^ of the Solochrome Black solution carefully. Shake to
mix and dilute to the 100-cm^ mark with water. Measure the absorbance
immediately at 520 nm against that of a blank solution, similarly prepared but
containing no magnesium.

XVIII, 24. MANGANESE. Discussion. Small quantities of manganese are
usually determined colorimetrically by oxidation to permanganic acid. The two
oxidising agents commonly used are ammonium persulphate in a phosphoric
acid-nitric acid medium in the presence of a little silver nitrate as catalyst, and
potassium periodate. The use of the latter will be described. In hot acid solution
(nitric or sulphuric acid), periodate oxidises manganese ion quantitatively to
permanganic acid:

2 Mn''-^ +5IO4- +3H2O = 2Mn04- +5IO3- + 6 H-"

The merits of the periodate method include: (n) the concentration of the acid
has little Influence, and may be varied within wide limits; {b) the boiling may be
prolonged beyond the time necessary to oxidise the manganese without
detriment; and (c) the permanganic acid solution will keep for several months
unchanged if an excess of periodate is present.

When ready for test the solution should not contain more than 2 mg of
manganese per 100 cm^, otherwise the colour will be too dark and colour
matching will be difficult. Frequently iron(III) is added to the standard in amount
equal to that found independently to be present in the sample. Phosphoric acid
must be present to prevent the precipitation of iron(III) periodate and iodate, and
also to decolorise the iron(III) (by complex formation). If chlorides are present it
is necessary to evaporate with a mixture of nitric and sulphuric acids until fumes
of the latter appear; chlorides react with the periodate. Reducing substances
reacting with periodate or permanganate must be destroyed (e.g., by evaporation
with nitric acid or a mixture of nitric acid and sulphuric acid) before the periodate
oxidation. Chromium(III) and cerium(III) are oxidised by periodate in acid
solution.

Procedure. For practice in this determination, the manganese content of a
standard steel sample may be evaluated. Weigh out accurately a suitable quantity
of the steel (0.1-0.2 g for steels containing up to 1 per cent of Mn) into a conical
flask, dissolve it in 20-50 cm^ of 1 : 3 nitric acid, and boil for 1 or 2 minutes to
expel oxides of nitrogen. Remove from the burner, and add 0.5-1.0 g A.R.
ammonium persulphate; boil for 10-15 minutes to oxidise carbon compounds
and to destroy the excess of persulphate. If any permanganate colour develops or
oxides of manganese separate, add a few drops of sulphurous acid or sodium
sulphite solution to reduce the manganese and render the solution clear, and boil
for a few minutes to expel the excess of sulphur dioxide. Dilute the solution to ca.
100 cm^, add 5-10 cm^ A.R. syrupy phosphoric acid and 0.5 g potassium
periodate (1): boil for I minute and keep hot for 5-10 minutes. Cool the solution
and make up to 250 cm^ in a graduated flask and mix thoroughly. Match the
colour against standards.

A wavelength of 545 nm should be used with a spectrophotometer. For steels or
other alloys, an aliquot of the sample solution, not oxidised but otherwise treated

746

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 25/26

similarly to the sample, may be used in the reference cell of the instrument.
Sometimes the blank consists of a portion of the developed sample to which a
reducing agent (nitrite) has been added to remove the permanganate colour: this
is not applicable when titanium is present.

Prepare standard manganese solutions by any of the following methods :

(a) Use a steel of known manganese content, which has been treated like the
unknown sample (2).

(b) Dissolve a known weight of pure electrolytic manganese in dilute nitric acid,
boil out oxides of nitrogen, and dilute to give a 0.01 per cent solution. If pure
manganese is not .available, reduce standard potassium permanganate

. solution with a little sulphite after the addition of dilute sulphuric acid, and
remove the sulphur, dioxide by boiling. Dilute the manganese(II) solution
and oxidise it with potassium periodate in the same way as the unknown
solution.

Notes. 1. Each 0.1 g of Mn requires 1 g of potassium periodate.

2. Such standards, containing excess of periodate, are stable for 2-3 months.

XVin, 25. MOLYBDENUM. Discussion. Molybdenum may be deter-
mined colorimetrically by the thiocyanate-tin(II) chloride method (for details, see
Section VI, 14) or by the dithiol method described here.

Toluene-3,4-dithiol, usually called dithiol, yields a slightly soluble dark green
complex, (CH 3 -C 6 H 3 -S 2 ) 3 Mo(VI), with molybdenum(VI) in a mineral acid
medium, which can be extracted by organic solvents. The resulting green solution
is used for the colorimetric determination of molybdenum.

Dithiol is a less-selective reagent than thiocyanate for molybdenum. Tungsten
interferes most seriously, but does not do so in the presence of tartaric acid or
citric acid (see Section XVIII, 29). Tin does not interfere if the absorbance is read
at 680 nm. Strong oxidants oxidise the reagent; iron(III) salts should be reduced
with potassium iodide solution and the liberated iodine removed with
thiosulphate.

Procedure. Prepare the dithiol reagent by adding 0.1-0.2 g of dithiol to 100
cm^ of 0.25M-sodium hydroxide solution, followed by 0.5 cm^ of thioglycollic
acid (to inhibit oxidation of the reagent); keep at 5 °C and prepare fresh daily.

Add to the sample solution (containing 1-25 //g of Mo) 4 cm^ of 1 : 3 sulphuric
acid, 3 drops of 85 per cent phosphoric acid, and 0.5 g of citric acid. Dilute with
water to 20 cm^ and add 2 cm^ of dithiol solution. Allow to stand at room
temperature for 2 hours. Extract the molybdenum complex with 13-cm^ and 10-
cm portions respectively of redistilled butyl acetate, and make up to 25.0 cm^
with this solvent in a graduated flask; filter through glass wool if not entirely
clear. Determine the absorbance of the solution at 670 nm. Prepare a calibration
curve as detailed in Section VI, 14.

XVin, 26. NICKEL Discussion. When dimethylglyoxime is added to an
alkaline solution of a nickel salt which has been treated with an oxidising agent
(such as bromine), a red coloration is obtained. The red soluble complex contains
wckel in a higher oxidation state.* The nickel complex formed absorbs at about
445 nm provided absorbance readings are made within 10 minutes of mixing. The

This has been regarded as nickel(III) and also as nickel(lV) dimethylglyoximate.

747

xvni, 27 QUANTITATIVE INORGANIC ANALYSIS

dimethylglyoxime-oxidising agent method must be distinguished from the
divalent nickel-dimethylglyoxime procedure which yields a nickel(II)
dimethylglyoximate soluble in chloroform: full details of this solvent extraction
method are given in Section VI, 15.

Cobalt(II), gold(UI), and dichromate ions interfere under the conditions of the
test. Metals which precipitate in ammoniacal solution can be removed by double
precipitation, or by taking advantage of the solubility of nickel(II)
dimethylglyoxime in chloroform (the nickelflll) complex is insoluble, as is also
the brown cobalt dimethylglyoxime). Copper may accompany the nickel in the
extraction; most of the copper is removed from the chloroform extract when it is
shaken with dilute ammonia solution, whereas the nickel remains in the organic
solvent. The nickel(II) dimethylglyoxime in the chloroform layer may be
decomposed by shaking with dilute hydrochloric acid; most of the
dimethylglyoxime remains in the chloroform, the nickel is transferred to the
aqueous phase and may be determined colorimetrically. Citrate or tartrate may
be added to prevent the precipitation of iron, aluminium, etc. Much manganese
may interfere, but this is prevented by adding hydroxylaramonium chloride
which maintains the element in the divalent state.

Procedure (nickel in steel). Dissolve 0.50 g, accurately weighed,* of the steel
in 10 cm^ of warm 1 : 1 nitric acid, boil to expel oxides of nitrogen, cool, and make
up to 250 cm^ with water in a graduated flask. Mix well, and transfer 5 cm^ of the
solution to a 50-cm^ graduated flask. To 5 cm^ of this solution add 5 cm^ of 10 per
cent citric acid solution, neutralise with concentrated ammonia solution and add
a few drops in excess (pH > 7.5). Add 2 cm^ of a 1 per cent dimethylglyoxime
solution in ethanol (or more if copper or cobalt is present). Extract with three 3-
cm^ portions of chloroform, shaking for 30 seconds each time. Shake the
combined chloroform extracts with 6 cm^ of 0.5M-ammonia solution (1:30);
shake the ammonia washing with 2 cm^ of chloroform and add the latter to the
main chloroform extract. Return the nickel to the ionic state by shaking the
chloroform extract vigorously for 1 minute with two 5-cra^ portions of 0.5M-
hydrochloric acid. Transfer the hydrochloric acid solution to a 25-cm^ graduated
flask, dilute to about 20 cm^, add 1 cm^ of saturated bromine water, followed by 2
cm^ of concentrated ammonia solution. Cool below 30 °C if necessary, add 1 cm^
of 1 per cent dimethylglyoxime solution, and dilute to volume. Measure the
absorbance at 445 nm after 5 minutes. The standard solutions for the
construction of the calibration curve should contain approximately the same
concentration of iron (nickel-free) as the sample solution.

Prepare the standard nickel solution by dissolving 0.673 g pure ammonium
nickel sulphate in water and diluting to 1 dm^: 1 cm^ contains 0.1 mg of Ni.
The solution may be further diluted to a basis of 0.01 mg of Ni per cm^, if
necessary. Pure nickel metal may also be employed for the preparation of the
standard solution.

XVIII, 27. TIN. Discussion. In acid solution, toluene-3, 4-dithiol (dithiol)
forms a red compound when warmed with tin(II) salts (compare Molybdenum,

* The weight of steel to be taken will naturally depend upon the nickel content. The final nickel
concentration should not exceed 0.6 mg per 100 cm^ because a precipitate may form above this
concentration.

748

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 11

Section XVin, 25). Tin(lV) also reacts, but more slowly than tin(II); thiogly collie
acid may be employed to reduce tin(IV) to tin(II). The reagent is not stable, being
easily reduced, and hence should be prepared as required. A dispersant is
generally added to the solution under test.

Many heavy metals react with dithiol to give coloured precipitates, e.g.,
bismuth, iron(III), copper, nickel, cobalt, silver, mercury, lead, cadmium, arsenic,
etc.; molybdate and tungstate also react. Of the various interfering elements,
only arsenic distils over with the tin when a mixture is distilled from a medium of
concentrated sulphuric acid and concentrated hydrobromic acid in a current of
carbon dioxide. If arsenic is present in quantities larger than that of the tin it
should be removed. ■

Reagents. Dithiol reagent. Dissolve 0.1 g dithiol in 2.5 cm^ 5M-sodium
hydroxide solution. Add 0.5 cm,^ thioglycollic acid, and dilute to 50 cm^ with
water. Prepare fresh daily.

Dispersant solution. Prepare a 1 per cent aqueous solution of sodium lauryl
sulphate.

Standard tin solution. Dissolve 1.000 g A.R. tin in 100 cm^ of 1 : 1
hydrochloric acid and dilute with the same concentration of acid to 1 dm^ : 1 cm^
contains 1 mg Sn. Prepare more dilute solutions as required (e.g., 0.01 mg Sn per
cm^) by dilution with 1 : 1 hydrochloric'acid.

Procedure. Transfer a 10-cm^ aliquot of the sample solution, which should
be 0.5M in hydrochloric acid and contain not more than 0.25 mg of tin, to a 25-
cm^ graduated flask, and add in the order given 1 drop thioglycollic acid, 2.0 cm^
concentrated hydrochloric acid, 0.5 cm^ of the dispersant solution, and 1.0 cm^ of
the dithiol reagent with thorough mixing after each addition. Place the flask in a
water bath at 60 °C for 10 minutes, cool, and dilute the contents to the mark.
Measure the absorbance at 545 nm against a reagent blank.

Construct a concentration-absorbance curve with the aid of the standard tin
solution.

Procedure (tin in canned foods). The procedure provides for the removal of
interfering copper by the addition of diethylammonium diethyldithiocarbamate
in chloroform reagent.*

Weigh 5 or 10 g of the sample, depending on the expected tin content, into a
small porcelain crucible. Dry and char the sample on a hot plate; heat to ash in a
muffle furnace at 600 °C. Add 1 g of fusion mixture (3 parts Na 2 C 03 + 1 part
KCN by weight) and fuse this with the ash by holding the crucible with nickel
tongs over a Bunsen or Meker burner. Cool the crucible, place it in a small
beaker, and cover the latter with a watch glass. Add 10 cm^ water, and run 10 cm^
dilute hydrochloric acid (1:1) cautiously into the crucible (Fume Cupboard!).
Boil the contents of the beaker gently for 30 minutes. Cool and filter: wash the
beaker, crucible, and filter with water.

If copper is known to be absent or present only in negligible proportions, dilute
the solution with water to 50 cm^ in a graduated flask, and continue.as detailed

elow. Otherwise, transfer the solution to a small separatory funnel and add 5
m diethylammonium diethyldithiocarbamate in chloroform reagent

chloroform when required). Shake and run off the
c oroform layer, extract the aqueous layer with successive 1-cm^ portions of the

This reagent K prepared from 3.0 cm^ of diethylamine in chloroform and 1 cm^ of carbon disulphide

cm of chloroform. Mix carefully and store in a dark bottle in a refrigerator.

749

XVIII, 28 QUANTITATIVE INORGANIC AN ALY SIS

reagent until the chloroform layer is colourless; finally, wash the aqueous layer
with a few cm^ of chloroform. Dilute the aqueous solution with water to 50 cm^ in
a graduated flask.

To 10.0 cm^ of the solution thus prepared add 0.5 cm^ of dilute hydrochloric
acid (1 : 1) and proceed as above. Measure the absorbance at 545 nm, or use an
Ilford No. 604 green filter with an absorptiometer.

XVIII, 28. TITANIUM. Discussion. With an acidic titanium(IV) solution
hydrogen peroxide produces a yellow colour;* with small amounts of titanium
(up to 0.5 mg of TiOj per cm^), the intensity of the colour is proportional to the
amount of the element present. Comparison is usually made with standard
titanium(IV) sulphate solutions; a method for their preparation from potassium
titanyl oxalate is described below. The hydrogen peroxide solution should be
about 3 per cent strength (10-volume) and the final solution should contain
sulphuric acid having a concentration from about 0.75 to 1.75M in order to
prevent hydrolysis to a basic sulphate and to prevent condensation to metatitanic
acid. The colour intensity increases slightly with rise of temperature, hence
the solutions to be compared should have the same temperature, preferably
20-25 °C.

Elements which interfere are: (a) iron, nickel, chromium, etc., because of the
colour of their solutions; (h) vanadium, molybdenum, and, under some
conditions, chromium, because they form coloured compounds with hydrogen
peroxide; (c) fluorine (even in minute amount) and large quantities of phosphates,
sulphates, and alkali salts (the influence of the last two is largely reduced the
greater the concentration of sulphuric acid present — up to 10 per cent). The
influence of elements of class (a) is overcome, if present in small amount, by
matching the colour by the addition of like quantities of the coloured elements to
the standard before hydrogen peroxide is added. When large amounts of iron are
present, as in the analysis of cast irons and steels, two methods may be adopted : (i)
phosphoric acid can be added in like amount to both unknown and standard,
after the addition of hydrogen peroxide; (ii) the iron content of the unknown
solution is determined, and a quantity of standard iron(IIl) alum solution,
containing the same amount of iron, is added to the standard solution. Large
quantities of nickel, chromium, etc., must be removed. Elements of class (b) must
also be removed; vanadium and molybdenum are most easily separated by
precipitation of the titanium with sodium hydroxide solution in the presence of a
little iron. Fluoride has the most powerful effect in bleaching the colour; it must
be removed by repeated evaporation with concentrated sulphuric acid. The
bleaching effect of phosphoric acid is overcome by adding a like amount to the
standard, or by adding 1 cm^ of 0. 1 per cent uranyl acetate solution for each 0.1
mg of Ti present.

Preparation of standard titanium solution. Weigh out 3.68 g A.R. potassium
titanyl oxalate K 2 TiO(C 2 04 ) 2,21120 into a Kjeldahl flask, add 8 g ammonium
sulphate and 100 cm^ concentrated sulphuric acid. Gradually heat the mixture to
boiling and boil for 10 minutes. Cool, pour the solution into 750 cm^ of water,
and dilute to 1 dm^ in a graduated flask; 1 cm^ = 0.50 mg of Ti.

* The coloured species formed has been stated to be [TiOfSO^lj]^ ~ or a similar ion; and it has also

been formulated as fri(H202)|‘** or an analogouscomplex.

750

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 29

If there is any doubt concerning the purity of the A.R. salt; standardise the
solution by precipitating the titanium with ammonia solution or with cupferron
solution, and ignite the precipitate to TiOj.

■Procedure^ The sample solution should preferably contain titanium as
sulphate in sulphuric acid solution, and be free from the interfering constituents
mentioned in the Discussion above. The final acidity may vary from 0.75 to
1.75M. If iron is present in appreciable amounts, add dilute phosphoric acid from
a burette until the yellow colour of the iron(III) is eliminated : the same amount of
phosphoric acid must be added to the standards. If alkali sulphates are present in
the test solution in appreciable quantity, add a like amount to the standards. Add
10 cm^ of 3 per cent hydrogen peroxide solution and dilute the solution to 100
cm^ in a graduated flask: the final concentration of Ti may conveniently be 2-25
parts per million. Compare the colour produced by the unknown solution with
that of standards of similar composition by any of the usual methods.

For a filter colorimeter use a blue filter (maximum transmission 400-420 nm);
a wavelength of 410 nm is employed for a spectrophotometer. In the latter case,
the effect of iron, nickel, chromium(III), and other coloured ions not reacting with
hydrogen peroxide may be compensated by using a solution of the sample, not
treated with hydrogen peroxide, in the reference cell.

XVni, 29. TUNGSTEN. Discussion. Toluene-3,4-dithiol (dithiol) may be
used for the colorimetric determination of tungsten; it forms a slightly soluble
coloured complex with tungsten(VI) which can be extracted with butyl or pentyl
acetate and other organic solvents. Molybdenum reacts similarly (see Section
XVni, 25) and must be removed before tungsten can be determined. The
molybdenum complex can be preferentially developed in cold weak acid solution
and selectively extracted with pentyl acetate before developing the tungsten
colour in a hot solution of increased acidity. The procedure will be illustrated by
describing the determination of tungsten in steel.

Reagents. Dithiol reagent solution. Dissolve 1 g toluene-3,4-dithiol in 100
cm^ pentyl acetate. This should be prepared immediately before use.

Standard tungsten solution. Dissolve 0.1794 g A.R. sodium tungstate
Na 2 W 04 , 2 H 20 in water and dilute to 1 dm^: 1 cm^ = 0.1 mg W. For use, dilute
100 cm^ of this solution to 1 dm^ : 1 cm^ = 0.01 mg W.

'Mixed acid’. Mix 15.0 cm^ concentrated sulphuric acid and 15.0 cm^
orthophosphoric acid (sp. gr. 1.75), and dilute to 100 cm^ with distilled water.

Procedure (tungsten in steel). Dissolve 0.5 g of the steel, accurately weighed,
in 30 cm^ of the ‘mixed acid’ by heating, oxidise with concentrated nitric acid, and
evaporate to fuming. Extract with 100 cm^ water, boil, transfer to a 500-cm^
graduated flask, cool, dilute to the mark with water, and mix. Pipette a.l5-cm^
aliquot into a 50-cm^ flask, evaporate to fuming, cool, add 5 cm^ dilute
hydrochloric acid (sp. gr. 1.06), warm until the salts dissolve, and cool to room
emperature. Add 5 drops of 10 per cent aqueous hydroxylammonium sulphate
20 dithiol reagent, and allow to stand in a bath at

U-25 Cfor 15 minutes withperiodicshaking.Transferthecontentsquantitatively
■° ®®P^^^tory funnel, using 3-4 cra^ portions of pentyl acetate for wash-

ing. Shake and allow the layers to separate. Run off the lower acid layer containing
e tungsten and reserve it in the original 50-cm^ flask. Wash the pentyl acetate
ayer twice consecutively with 5-cm^ portions of hydrochloric acid (sp. gr. 1.06)
n combine the acid washings with the original acid layer. Discard the

751

XVni, 30 QUANTITATIVE INORGANIC ANALYSIS

molybdenum-containing pentyl acetate layer. Evaporate the acid tungsten
solution carefully to fuming (to expel dissolved pentyl acetate), then add a few
drops of concentrated nitric acid during fuming to clear up any charred organic
matter. Add 5 cm^ of 10 per cent tin(II) chloride solution (in concentrated
hydrochloric acid) and heat to 100 °C for 4 minutes: add 10 cm^ of the dithiol
reagent and heat at 100 °C for 10 minutes longer with periodic shaking. Transfer
to a 25-cm^ stoppered separatory funnel, and rinse thrice with 2 -cm^ portions of
pentyl acetate. Shake, separate, and draw off the lower acid layer and discard it.
Add 5 cm^ concentrated hydrochloric acid to the organic layer, repeat the
extraction and again discard the lower layer. Draw off the pentyl acetate layer
containing the tungsten complex into a 50-cm^ graduated flask and dilute to
volume with pentyl acetate. Measure the absorbance with a spectrophotometer
at 630 nm in 4-cm cells, or use an absorptiometer and an Ilford Spectrum Red No.
608 filter.

Refer the readings to a calibration curve prepared from spectrographically
pure iron to which suitable amounts of standard sodium tungstate solution have
been added.

XVni, 30. VANADIUM. Of the two methods commonly used for the
determination of vanadium the second, in which phosphotungstovanadic acid is
formed, is employed most frequently.

A. Vanadyl Sulphate method. Discussion. When hydrogen peroxide is
added to a solution containing small quantities of vanadium(V) (up to 0.1 mg of V
per cm^) in sulphuric acid solution, a reddish-brown coloration is produced; this
is thought to be due to the formation of a compound of the type (^ 0 ) 2 ( 804 ) 3 . A
large excess of hydrogen peroxide tends to reduce the colour intensity and to
change the colour from red-brown to yellow. With a hydrogen peroxide
concentration of 0.03 per cent, the sulphuric acid concentration can vary between
0.3 and 3AI without any appreciable effect on the colour: with higher
concentrations of hydrogen peroxide, the acidity must be increased to permit
development of the maximum colour intensity.

The colour is unaffected by the presence of phosphate or fluoride. Titanium
and molybdenum( VI) (which give colours with hydrogen peroxide) and tungsten
interfere. Titanium may be removed by adding fluoride or hydrofluoric acid,
which simultaneously remove the yellow colour due to iron(III). If titanium is
absent, phosphate may be used to decolorise any iron(III) salt present. Oxalic
acid eliminates the interference due to tungsten. In the presence of elements which
yield coloured solutions, such as chromium or nickel, it is best to add equal
amounts of these elements to the standard solution. If steel is being analysed, the
most convenient procedure is to use a like steel as standard.

Prepare a standard vanadium solution by dissolving 1.146 g A.R. ammonium
vanadate in water and make up to 1 dm^; 1 cm^ = 0.5 mg of V. The above
solution may be diluted further to give a solution containing 0.01 mg V per cm^.

Procedure. Make the solution 0.5-liVf in sulphuric acid and add 0.25 cm^
of 3 per cent hydrogen peroxide for each 10 cm^ of test solution. Compare
colorimetrically against a standard having the same acidity and containing the
same volume of hydrogen peroxide solution. If titanium is present, add
hydrofluoric acid (say, 5-10 per cent of the volume); this will also decolorise the
iron(III). If titanium is absent, use phosphoric acid for the decolorisation of the
iron.

752

COLORIMETRY AND SPECTROPHOTOMETRY XVffl, 31

The absorbance due to the vanadyl sulphate may be measured at 450 nm (or at
290 nm in the ultraviolet) against a reagent blank or compensating blank.

B. Phosphotungstovanadic acid method. Discussion. Vanadium may also
be determined by making use of the yellow, soluble phosphotungstovanadic acid
formed upon adding phosphoric acid and sodium tungstate to an acid vanadate
solution. The most intense colour is obtained when the molecular ratio of
phosphoric acid to sodium tungstate is in the range 3 : 1 to 20 : 1, and the tungstate
concentration in the test solution isO.OltoO.lM; the preferred concentrations are
0.5M in phosphoric acid and 0.025M in sodium tungstate.

The following interfere; (a) coloured ions, such as chromate, copper, and
cobalt; (b) titanium, zirconium, bismuth, antimony, and tin yield slightly soluble
phosphates or basic salts except in very low concentrations; (c) potassium and
ammonium ions give sparingly soluble phosphotungstates; (d) molybdenum(VI)
in relatively high concentration (>0.5 mg cm“^); (e) iodide, thiocyanate, etc.,
reduce phosphotungstic acid; and (/) iron in concentration greater than 1 mg
cm" ^ (slight interference even in the presence of phosphoric acid).

Procedure. Render the solution ca. 0.5 M in mineral acid, and add 1.0 cm^
of 1 ; 2 phosphoric acid and 0.5 cm^ of 0.5M-sodium tungstate solution (prepared
by dissolving 16.5 g A.R. sodium tungstate Na 2 W 04 , 2 H 20 in 100 cm^ water) for
each 10 cm^ of test solution. Heat to boiling, cool, dilute to volume, and
determine the absorbance of the resulting solution* at 400 nm. If small amounts
of coloured ions (nickel, cobalt, dichromate, etc.) are present, these should be
incorporated in the comparison solution, preferably by employing an aliquot
portion of the original sample solution.

Vanadium in steel. Dissolve 1.0 g, accurately weighed, of the steel in 50 cm^
of 1:4 sulphuric acid. When solution is complete, introduce 10 cm^ of
concentrated nitric acid, and boil until nitrous fumes are no longer evolved.
Dilute the solution to 100 cm^ with hot water, heat to boiling, and add saturated
potassium permanganate solution until a pink colour persists or a precipitate is
formed. Boil for 5 minutes. Filter off any tungsten(VI) oxide or manganese oxide
which may be precipitated. Add a slight excess of freshly prepared sulphurous
acid, and boil off the excess. Cool, add 5 cm^ syrupy phosphoric acid and 5 cm^ of
10-volume hydrogen peroxide.

Simultaneously with the main determination prepare in an analogous manner
a comparison solution from a standard steel which contains no vanadium but is
otherwise similar; add a standard solution of vanadium to the control, followed
oy hydrogen peroxide, etc., and compare this colorimetrically or
spectrophotometrically with the solution obtained from the unknown steel.

Anions

Xyni, 32 _ CHLORIDE. Two procedures are commonly employed for the
colorimetric determination of chloride.

chi j^®rcury(II) chloranilate method. Discussion. The mercury(II) salt of
oramlic acid (2,5-dichloro-3,6-dihydroxy-p-benzoquinone) may be used for

‘The yellow colour
reagent blank.

may also be extracted with 2-methylpropanol,

and read at 400 nm against a

753

XVIII, 31 QUANTITATIVE INORGANIC ANALYSIS

the determination of small amounts of chloride ion. The reaction is:

HgQCl 204 + 2 Cr + H+ = HgClz + HCfiCljO^-

The amount of reddish-purple acid-chloranilate ion liberated is proportional to
the chloride-ion concentration. Methyl cellosol ve (2-methoxyethanol) is added to
lower the solubility of niercury(n) chloranilate and to suppress the dissociation of
the mercury(II) chloride; nitric acid is added (concentration 0.05M) to give the
maximum absorption. Measurements are made at 530 nm in the visible or 305 nm
in the ultraviolet region. Bromide, iodide, iodate, thiocyanate, fluoride, and
phosphate interfere, but sulphate, acetate, oxalate, and citrate have little effect at
the 25-p.p.m. level. The limit of detection is 0.2 p.p.m. of chloride ion; the upper
limit is about 120 p.p.m. Most cations, but not ammonium ion, interfere and must
be removed.

Silver chloranilate cannot be used in the determination because it produces
colloidal silver chloride.

Procedure. Remove interfering cations by passing the aqueous solution
containing the chloride ion through a strongly acidic ion exchange resin in the
hydrogen form (e.g., Zerolit 225 or Amberlite 120) contained in a tube 15 cm
long and 1.5 cm in diameter. Adjust the pH of the effluent to 7 with dilute nitric
acid or aqueous ammonia and pH paper. To an aliquot containing not more than
1 mg of chloride ion in less than 45 cm^ of water in a 100-cm^ graduated flask, add
5 cm^ IM-nitric acid and 50 cm^ methyl cellosolve. Dilute the mixture to volume
with distilled water, add 0.2 g mercury(II) chloranilate, and shake the flask
intermittently for 15 minutes. Separate the excess of mercury(II) chloranilate by
filtration through a fine ashless filter paper or by centrifugation. Measure the
absorbance of the clear solution with a spectrophotometer at 530 nm against a
blank prepared in the same manner.

Construct a calibration curve using standard ammonium chloride solution
(1-100 p.p.m. Cl") and deduce the chloride-ion concentration of the test solution
with its aid.

Mercury(II) chloranilate may be prepared by adding dropwise a 5 per cent
solution of A.R. mercury(II) nitrate in 2 per cent nitric acid to a stirred solu-
tion of chloranilic acid at 50 °C until no further precipitate forms. Decant the
supernatant liquid, wash the precipitate thrice by decantation with ethanol, once
with diethyl ether, and dry in a vacuum oven at 60 °C. The compound is
available commercially.

B. Mercury(II) thiocyanate method. This second procedure for the
determination of trace amounts of chloride ion depends upon the displacement of
thiocyanate ion from mercury(II) thiocyanate by chloride ion; in the presence of
iron(III) ion with a highly coloured iron(III) thiocyanate complex is formed, and the
intensity of its colour is proportional to the original chloride-ion concentration:

2C1 - + Hg(SCN)2 + 2Fe^ = HgCl^ + 2[Fe(SCN)]‘ ^

The method is applicable to the range 0.5-100 //g of chloride ion.

Procedure. Place a 20-cm^ aliquot of the chloride solution in a 25-cm^
graduated flask, add 2.0 cm^ of 0.25M-ammonium iron(III) sulphate {Fe(NH4)
(804)2, 12H2O} in 9M-nitric acid, followed by 2.0 cm^ of a saturated solution of
mercury(II) thiocyanate in ethanol. After 10 minutes measure the absorbance of
the sample solution and also of the blank in 5-cm cells in a spectrophotometer at
460 nm against water in the reference cell. The amount of chloride ion in the

754

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 32/33

sample corresponds to the difference between the two absorbances and is
obtained from a calibration curve.

Construct a calibration curve using a standard sodium chloride solution
containing 10 jUg Cl “ per cm^ : cover the range 0-50 jig as above. Plot absorbance
against micrograms of chloride ion.

XVin,32. FLUORIDE. Fluoride, in the absence of interfering anions
(including phosphate, molybdate, citrate, and tartrate) and interfering cations
(including cadmium, tin, strontium, iron, and particularly zirconium, cobalt, lead,
nickel, zinc, copper, and aluminium), may be determined with thorium
chloranilate in aqueous 2-methoxyethanol at pH 4.5 ; the absorbance is measured
at 540 nm or, for small concentrations 0-2.0 p.p.m., at 330 nm.

In water as solvent, the reaction is:

TKCfiClaOJ^ + dF- +2H+ ^ThF^^- + 2 HC 6 CI 2 O 4 -
In aqueous 2 -methoxyethanol, the main reaction is stated to be:

Th(C6Cl204)2 + 2F- +H+ ^ThF2C6Cl204 + HC6Cl204-

Interfering cations, except aluminium and zirconium, can be removed by
passage through an ion exchange column! In the presence of interfering anions
and also aluminium and zirconium, fluoride may be separated as hydrofluosilicic
acid by distilling with dilute perchloric acid at 135 °C (temperature maintained by
the addition of water) in the presence of a few glass beads.

A calibration curve for the range 0.2-10 mg fluoride ion per 100 cm^ is
constructed as follows. Add the appropriate amount of standard sodium fluoride
solution, 25 cm^ of 2-methoxyethanol, and 10 cm^ of a buffer (O.IM in both
sodium acetate and acetic acid) to a 100-cm^ graduated flask. Dilute to volume
with distilled water and add about 0.05 g of thorium chloranilate. Shake the flask
intermittently for 30 minutes (the reaction in the presence of 2-methoxyethanol is
about 90 per cent complete after 30 minutes and almost complete after 1 hour)
and filter about 10 cm^ of the solution through a dry Whatman No. 42 filter
paper. Measure the absorbance of the filtrate in a 1 -cm cell at 540 nm against a
blank, prepared in the same manner, using a suitable spectrophotometer. Prepare
a calibration curve for the concentration range 0 . 0 - 0.2 mg fluoride ion per 100
cm in the same way, but add only 10.0 cm^ of 2 -methoxyethanol; measure the
absorbance of the filtrate in a 1 -cm silica cell at 330 nm.

Treat the fluoride sample solution in the same manner as described for the
calibration curve after removing interfering ions and adjusting the pH to about 5
with dilute nitric acid or sodium hydroxide solution. Read off the fluoride
concentration from the calibration curve and the observed value of the
absorbance.

> 33. NITRITE. Discussion. General procedures for the deter-
^ ation of nitrites are usually based upon some form of diazotisation reaction,
en involving carcinogenic materials such as the naphthylamines. In the
oiiowmg method these compounds are avoided.

suloh 1 nitrite ion, under acidic, conditions, causes diazotisation of

counl^H* (4-aminobenzenesuiphonamide) to occur, and the product is
pc With lV-(l-naphthyl)-ethylenediamine dihydrochloride.

10ft ^^hhanilamide solution {A). Dissolve 0.5 g.sulphanilamide in

of 20 per cent v/v hydrochloric acid.

755

XVIII, 34 QUANTITATIVE INORGANIC ANALYSIS >

'N-{l-napIithy[)-ethyle?iediamine dihydrochloride solution (B). Dissolve 0.3 g of
the solid reagent in 100 cm^ of 1 per cent v/v hydrochloric acid.

. Procedure. To 100 cm^ of the neutral sample solution (containing not more
than 0.4 mg nitrite) add 2.0 cm^ of solution A and, after 5 minutes, 2.0 cm^ of
solution B. The pH at this point should be about 1.5. Measure the absorbance
after 10 minutes in the wavelength region of 550 nm in a spectrophotometer
against a blank solution prepared in the same manner. Calculate the
concentration of the nitrite from a calibration plot prepared from a series of
standard nitrite solutions.

XVin, 34. PHOSPHATE. Two methods are commonly used for the
determination of phosphate.

A. Molybdenum Blue method. Discussion. Orthophosphate and molyb-
date ions condense in acidic solution to give molybdophosphoric acid
(phosphomolybdic acid), which upon selective reduction (say, with hydrazinium
sulphate) produces a blue colour, due to molybdenum blue of uncertain
composition. The intensity of the blue colour is proportional to the amount of
phosphate initially incorporated in the heteropoly acid. If the acidity at the time
of reduction is 0.5M in sulphuric acid and hydrazinium sulphate is the reductant,
the resulting blue complex exhibits maximum absorption at 820-830 nm.

Ions which form heteropoly acids, such as silicate (Section XVIII, 35), arsenate
(Section XVIII, 14), germanale, and tungstate, should be absent. Silicate may be
separated by fuming with perchloric acid to dehydrate the silicic acid and render
it insoluble. Arsenate can be volatilised as arsenic(III) bromide from a
hydrobromic acid-sulphuric acid solution; tin and germanium are volatilised
simultaneously. Lead, antimony, and copper interfere; these and other metals
may be removed by passage of the solution through a cation-exchange column.
Oxidising and reducing agents must be absent.

Reagents. Molybdate solution. Dissolve 1 2.5 g of A.R. sodium molybdate
(Na 2 Mo 04 , 2 H 20 ) in 5M-sulphuric acid and dilute to 500 cm^ with 5M-
sulphuric acid.

Hydrazinium sulphate solution. Dissolve 1.5 g of A.R. hydrazinium
sulphate in de-ionised water and dilute to 1 dm^.

Standard phosphate solution. Dissolve 0.2197 g of A.R. potassium
dihydrogen phosphate in de-ionised water and dilute to 1 dm^ in a graduated
flask. 1 cm^ = 0.05 mg P. Dilute as appropriate.

Procedure. The sample solution should contain not more than 0.1 mg of
phosphorus as the orthophosphate in 25 cm^ and should be neutral. Transfer 25
cm^ to a 50-cm^ Pyrex graduated flask. Add 5.0 cm^ of the molybdate solution,
followed by 2.0 cm^ of the hydrazinium sulphate solution, dilute to the mark with
distilled water, and mix well. Immerse the flask in a boiling water bath for 10
minutes, remove, and cool rapidly. Shake the flask, adjust the volume, and
measure the absorbance at 830 nm against either de-ionised water or a reagent
blank.

Construct the calibration curve, using the standard phosphate solution, in the
usual manner.

B. Phosphovanadomolybdate method. Discussion. This second method
(Ref. 10) is considered to be slightly less sensitive than the previous Molybdenum
Blue method, but it has been particularly useful for phosphorus determinations

756

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 35

carried out by means of the Oxygen (Schoniger) Flask (Section in, 46). The
phosphovanadomolybdate complex formed between the phosphate, ammonium
vanadate, and ammonium molybdate is bright yellow in colour and its
absorbance can be measured between 460 and 480 nm.

Reagents. Ammonium vanadate solution. Dissolve 2.5 g ammonium
vanadate (NH 4 VO 3 ) in 500 cm^ hot water, add 20 cm^ concentrated nitric acid
and dilute with water to 1 dm^ in a graduated flask.

Ammonium molybdate solution. Dissolve 50 g ammonium molybdate,
(NH 4 ) 6 Mo 7024 - 4 H 20 , in warm water and dilute to 1 dm^ in a graduated flask.
Filter the solution before use.

Procedure. Dissolve 0.4 g of the phosphate sample in 2.5M-nitric acid to
give 1 dm^ in a graduated flask. Place a 10-cm^ aliquot of this solution in a 100-
cm^ graduated flask, add 50 cm^ water, 10 cm^ of the ammonium vanadate
solution, 10 cm^ of the ammonium molybdate solution and dilute to the mark.
Determine the absorbance of this solution at 465 nm against a blank prepared in
the same manner, using 1 cm cells.

Prepare a series of standards from potassium dihydrogenphosphate covering
the range 0-2 mg phosphorus per 100 cm^ and containing the same
concentration of acid, ammonium vanadate, and ammonium molybdate as the
previous solution. Construct a calibration curve and use it to calculate the
concentration of phosphorus in the sample.

XVni, 35. SILICATE. Discussion. Small quantities of dissolved silicic acid
react with a solution of a molybdate in an acid medium to give an intense yellow
coloration, due probably to the complex molybdosilicic acid H 4 [SiMoi 204 o].
T^e latter may be employed as a basis for the colorimetric determination of
silicate (absorbance measurements at 400 nm). It is usually better to reduce the
complex acid to molybdenum blue (the composition is uncertain); a solution of
a mixture of l-amino-2-naphthol-4-sulphonic acid and sodium hydrogen
sulphite solution is satisfactory reducing agents.

Phosphates, arsenates, and germanates give similar colorations and must
either be removed or their interferences eliminated by the addition of suitable
reagents: arsenic and germanium can be removed by evaporation with
nydrochloric acid, and phosphate by precipitation as ammonium magnesium
phosphate in acetic acid solution, or may be rendered innocuous by the addition
0 ammonium citrate. Elements such as barium, bismuth, lead, and antimony give
precipitates or turbidities, and must be absent. Water used for dilution should be
res ly distilled in an all-Pyrex apparatus or passed through a mixed-bed ion
exc ange column, and stored in polythene containers. Water tends to dissolve
sioUi cant traces of silica on standing in glass, particularly soda glass, vessels.

I Ammonium molybdate solution. Dissolve 8.0 g A.R. ammonium

ion ^ 3 ^*^ ^’^ystals in water, add 9 cm^ concentrated sulphuric acid, and dilute to

cm^ w^t ^Sfnt. Solution A: dissolve 10 g sodium hydrogensulphite in 70
cm^ Solution B; dissolve 0.8 g anhydrous sodium hydrogensulphite in 20

witVi S l-amino-2-naphthol-4-sulphonic acid. Mix solution A

with ^olution B, and dilute to 100 cm^

^ artaric ac/d solution. Prepare a 10 per cent aqueous solution.

1-0 R oil R™ oj silica. Fuse 0.107 g of pure, dry precipitated silica with

- anhydrous sodium carbonate in a platinum crucible. Cool the melt.

757

XVin, 36 QUANTITATIVE INORGANIC ANALYSIS

dissolve it in de-ionised water, dilute to 500 cm^, and store in a polythene bottle.
1 cm^ = 0.1 mg Si. Dilute as appropriate, say, to 1 cm^ = 0.01 mg Si.

Procedure. The sample solution, free from interfering elements and
radicals, may conveniently occupy a volume of about 50 cm^ and contain
between 0.01 and 0.1 mg of silica; the pH should be 4.5-5.0. Add 1 cm^ of the
ammonium molybdate solution and, after 5 minutes, add 5 cm^ of the tartaric
acid solution and mix. Introduce 1.0 cm^ of the reducing agent and dilute to 100
cm^ in a graduated flask. Measure the absorbance at ca. 815 nm after 20 minutes
against de-ionised water.

Construct a calibration curve using 0, 1.0, 2.5, 5.0, 7.5, and 10.0 cm^ of the
standard silica solution (1 cm^ s 0.01 mg Si) which have been treated similarly.

XVm, 36. SULPHATE. Discussion. The barium salt of chloranilic acid
(2,5-dichloro-3,6-dihydroxy-p-benzoquinone) illustrates the principle of a
method which may find wide application in the colorimetric determination of
various anions. In the reaction

Y” + MA (solid) = A~ + MY (solid)

where Y" is the anion to be determined and A” is the coloured anion of an
organic acid, MY must be so much less soluble than MA that the reaction is
quantitative. M A must be only sparingly soluble so that the blanks will not be too
high. Sulphate ion in the range 2-400 p.p.m. may be readily determined by
utilising the reaction between barium chloranilate with sulphate ion in acid
solution to give barium sulphate and the acid-chloranilate ion:

SO.;^- -i-BaCfiCljO^-b H"- = BaSO^-l-HCsClzO^-

The amount of acid chloranilate ion liberated is proportional to the sulphate-ion
concentration. The reaction is carried out in 50 per cent aqueous ethanol,
buffered at an apparent pH of 4. Most cations must be removed because they
form insoluble chloranilates: this is simply effected by passage of the solution
through a strongly acidic ion-exchange resin in the hydrogen form (see Section
Vn, 2). Chloride, nitrate, hydrogencarbonate, phosphate, and oxalate do not
interfere at the 100-p.p.m. level. The pH of the solution governs the absorbance of
chloranilic acid solutions at a particular wavelength; chloranilic acid is yellow,
acid-chloranilate ion is dark purple, and chloranilate ion is light purple. At pH 4
the acid-chloranilate ion gives a broad peak at 530 nm, and this wavelength is
employed for measurements in the visible region. A much more intense
absorption occurs in the ultraviolet: a sharp band at 332 nm enables the limit of
detection of sulphate ion to be extended to 0.06 p.p.m.

Procedure. Pass the aqueous solution containing sulphate ion (2-400
p.p.m.) through a column 1.5 cm in diameter and 15 cm long of Zerolit 225 or
equivalent resin in the hydrogen form. Adjust the effluent to pH 4 with dilute
hydrochloric acid or ammonia solution. Make up to volume in a graduated flask.
To an aliquot containing up to 40 mg of sulphate ion in less than 40 cm^ in a 100-
cm^ graduated flask, add 10 cm^ of a buffer (pH = 4; a 0.05M solution of A.R.
potassium hydrogenphthalate) and 50 cm^ of 95 per cent ethanol. Dilute to the
mark with distilled water, add 0.3 g of barium chloranilate and shake the flask for
10 minutes. Remove the precipitated barium sulphate and the excess of barium
chloranilate by filtering or centrifuging. Measure the absorbance of the filtrate
with a filter colorimeter or a spectrophotometer at 530 nm against a blank

758

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 37

prepared in the same manner. Construct a calibration curve using standard
potassium sulphate solutions prepared from the A.R. salt.

Experimental. Determinations with ultraviolet/visible
spectrophotometers

XVin, 37. DETERMINATION OF THE ABSORPTION CURVE AND
CONCENTRATION OF A SUBSTANCE (POTASSIUM NITRATE). Dis-
cussion. Potassium nitrate is an. example of an inorganic compound which
absorbs mainly in the ultraviolet, and can be employed to obtain experience in
the use of a manually operated ultraviolet/visible spectrophotometer. Some of
the exercise can also be carried out employing an automatic recording
spectrophotometer. •

The absorbance and the percentage transmission of an approximately O.IM-
potassium nitrate solution is measured over the wavelength range 240-360 nm at
5-nm intervals and at smaller intervals in the vicinity of the maxima or minima.
Manual spectrophotometers are calibrated to read both absorbance and
percentage transmission on the dial settings, while the automatic recording
double beam spectrophotometers usually use chart paper printed with both
scales. The linear conversion chart. Fig. XVIII, 36, is useful for visualising the
relationship between these two quantities.

The three normal means of presenting the spectrophoto-
metric data are described below: by far the most common
procedure is to plot absorbance against wavelength (measured
in nanometres). The wavelength corresponding to the absorb-
ance maximum (or minimum transmission) is read fronj the
plot and is used for the preparation of the calibration curve.
This point is chosen for two reasons : (i) it is the region in which
the greatest difference in absorbance between any two different
concentrations will be obtained, thus giving the maximum
sensitivity for concentration studies, and (ii) as it is a turning
point on the curve it gives the least alteration in absorbance
value for any slight variation in wavelength.

No general rule can be given concerning the strength of the
solution to be prepared, as this will depend upon the
spectrophotometer used for the study. Usually a 0.01-0. OOlAf
solution is sufficiently concentrated for the highest absorb-
ances, and other concentrations are prepared by dilution. The
concentrations should be selected such that the absorbance lies
between about 0.3 and 1.5.

For the determination of the concentration of a substance,
select the wavelength of maximum absorption for the
compound (e.g., 302.5-305 nm for potassium nitrate) and
construct a calibration curve by measuring the absorbances of
four or five concentrations of the substance (e.g., 2, 4, 6 , 8 and
10 g KNO 3 dm“^) at the selected wavelength. Plot absorbance
(ordinates) against concentration (abiscissae). If the compound
%XVlli ^ssr’s law a linear calibration curve, passing through the

’ origin, will be obtained. If the absorbance of the unknown

759

xvm, 37 QUANTITATIVE INORGANIC ANALYSIS

solution is measured the concentration can be obtained from the calibration
curve.

If it is known that the compound obeys Beer’s law the molar absorption
coefficient e can be computed from one measurement of the absorbance of a
standard solution. The unknown concentration is then calculated using the value
of the constant e and the measured value of the absorbance under the same
conditions.

Procedure. Dry some A.R. potassium nitrate at 1 10°C for 2-3 hours and
cool in a desiccator. Prepare an aqueous solution containing 10.000 g dm“^.
With the aid of a precision spectrophotometer* and matched 10-mm rectangular
cells, measure the absorbance and the percentage transmission over a series of
wavelengths covering the range 240—350 nm. Plot the data in three different
ways: (i) absorbance against wavelength; (ii) percentage transmission against
wavelength; and (iii) log c (molecular decadic absorption coefficient) against
wavelength. The curves obtained for potassium nitrate are shown in Figs. XVIII,
37 to XVin, 39. From the curves, evaluate the wavelength of maximum
absorption (or minimum transmission). Use this value of the wavelength to
determine the absorbance of solutions of potassium nitrate containing 2.000,
4.000, 6.000, and 8.000 g of potassium nitrate dm“ Run a blank on the two cells,
filling both the blank cell and the sample cell svith distilled water; if the cells are
correctly matched no difference in absorbance should be discernible. Plot the
absorbances (ordinates) against concentration.

Determine the absorbance of an unknown solution of potassium nitrate and
read the concentration from the calibration curve.

When reporting spectrophotometric measurements, details should be given of the concentration
used, the solvent employed, the make and model of the instrument, as well as the slit widths
employed, together with any other pertinent information.

760

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 38

xvni, 38 . SPECTROPHOTOMETRIC DETERMINATION OF THE pK
VALUE OF AN INDICATOR (THE ACID DISSOCIATION CONSTANT
OF METHYL RED). Discussion. The dissociation of an acid-base indicator is
well suited to spectrophotometric study; the procedure involved will be
illustrated by the determination of the acid dissociation constant of methyl red
(MR). The acidic (HMR) and basic (MR “) forms of methyl red are shown below.

ACID FORM (HMR) RED

BASIC FORM (MR-) YELLOW

The acid dissociation constant K is given by the equation:
j._ [H"][MR-]

[HMR]

( 1 )

pK = pH-log M - ^ (2)

®[HMR] ^ ’

Both HMR and MR" have strong absorption peaks in the visible portion of the
spectrum; the colour change interval from pH 4 to pH 6 can be conveniently
obtained with a sodium acetate-acetic acid buffer system.

The determination of pK involves three steps :

(«) Evaluation of the wavelengths at which HMR (A^) and MR" (Ab) exhibit
maximum absorption.

ib) Verification of Beer’s law for both HMR and MR" at wavelengths A^ and Ag.
(c) Determination of the relative amounts of HMR and MR" present in
solution as a function of pH.

By using the same concentration of indicator in each of the measurements at
t of pH and measuring the absorbance for each solution at A^ and

^ B) the relative amounts of HMR and MR" in solution can be calculated from

the two equations:

'^A = dA.HMR[HMR]+d^^^

~ ^b.hmr[HMR]

where d.

[MR"]
B.MR- [MR ]

(3)

(4)

Bv > dB,HMR and (/b.mr- are derived from the graphs plotted in (h).

obtai simultaneous equations, the ratio [MR"]/[HMR] can be

that th^ thence pK with the aid of equation (2). Equations (3) and (4) imply

ahcnrF^ Observed absorbances {A) at A* and Ab are the simple additive sums of the
absorbances (d) due to HMR and MR".

ted in 30^^” 3 ^ Met/iy/ red solution. Dissolve 0.10 g pure crystalline methyl
cm 95 per cent ethanol and dilute to 50 cm^ with water. The solution

761

xvm, 38 QUANTITATIVE INORGANIC ANALYSIS

required in the experiment (standard solution) is prepared by transferring 5.0 cm^
of the above stock solution to 50 cm^ of 95 per cent ethanol contained in a 100-
cm^ graduated flask and diluting to 100 cm^ with water.

Sodium acetate, 0.04M and 0.01 M.

Acetic acid, 0.02M.

Hydrochloric acid, O.IM and O.OIM. The exact concentrations of these two
solutions are not critical.

Procedure. The study can be carried out using either a manually operated
single beam spectrophotometer, or an automatic recording double beam
spectrophotometer. In both cases the wavelengths at which HMR and MR'
exhibit absorption maxima are readily obtained from the spectra.

(a) Prepare solution A by diluting a mixture of 10.0 cm^ of the standard
solution of the indicator (MR) and 10.0 cm^ of O.lM-hydrochloric acid to 100
cm^ ; the pH of this solution is about 2, so that the indicator MR is present entirely
as HMR. Using 1-cm cells, determine the absorption spectrum of this solution
over the range 350-600 nm against a blank of distilled water. For manual plotting
cover the range in increments of 25 nm except for the portion between 500 and
550 nm which should be covered in 10 nm increments. From the spectrum of
absorbance against wavelength determine the wavelength at which the
maximum absorbance occurs: this is about 520 nm.

Prepare solution B by diluting a mixture of 10.0 cm^ of the standard solution of
the indicator and 25.0 cm^ of 0.04M-sodium acetate to 100 cm^. The pH of this
solution is about 8, so that the indicator MR is present entirely as MR". Measure
the absorbance of solution B over the range 350-600 nm as detailed for solution
A; with a manual spectrophotometer use 25-nm steps except for 400-450 nm,
where 10-nm steps are recommended. Determine the wavelength Ag of maximum
absorbance as above: this is about 430 nm. The type of plots obtained for
solutions A and B is shown in Fig. XVIII, 40. The absorption peaks are not
completely separated, but cross at a wavelength of about 460 nm. This point is
known as the isobestic point. If the absorbance of a solution containing both
HMR and MR" is measured at this wavelength, the observed absorbance is
independent of the relative amounts of HMR and MR ' present and depends only
on the total amount of the indicator MR in the solution.

Wavelength, nm Relative concentrations of MR

HMR.Xa

MR", Xg

HMR,Xb
MR-, Xa

Fig. xvm, 40

Fig. xvm, 41

(b) Using solution A, measure out 40.0 cm^ 25.0 cm^, and 10.0 cm^ into
separate 50-cm^ graduated flasks, and dilute in each case to the mark with O.OIM-

762

COLORIMETRY AND SPECTROPHOTOMETRY XVIII, 39

hydrochloric acid. The resulting solutions will contain 0.8, 0.5, and 0.2 times
respectively the initial concentration of HMR. Similarly, using solution B and
diluting with O.OlM-sodium acetate, prepare three solutions containing
respectively 0.8, 0.5, and 0.2 times the initial concentration of MR~. Measure the
absorbance of each of the six solutions versus water at wavelengths of 2^ and Ab . It
is important in obtaining the experimental absorbance to be sure that all the
measurements are made at constant temperature, say, at the temperature of the
room housing the spectrophotometer. Plot absorbance against relative
concentration of the indicator MR: in each case straight-line plots should be
obtained, as in Fig. XVIII, 41.

Flask number

Standard indicator solution MR (cm^)

10.0

0.04At-Sodium acetate (cm^)

25.0

0.02jM-Acetic acid (cm^)

25.0

10.0

5.0

Water (to mark)

15.0

55.0

60.0

4.84

5.15

5.53

5.81

Determine the pH values of each of the solutions (typical values are incorporated
in the table) and measure the absorbance of each solution at wavelengths and
-Ib' All these solutions contain the same concentration of indicator as solutions A
and B used in (a). For each prepared solution, obtain the values of the
absorbances ^umr, ^^a.mr- > 4.hmr, and dB.MR- from the plots in Fig. XVIII, 41, at
relative concentrations of 1.0, and solve the simultaneous equations (3) and (4) in
order to evaluate the relative amounts of HMR and MR" in solution. From the
relative amounts of HMR and MR" present as a function of pH, calculate the
value of pK for methyl red using equation (2). Some typical results are collected in
the following table:

Soludon

Observed

Absorbance

[MR-]

number

at^A

atAfl

[HMR]

4.84

0.605

0.204

0.679

5.01

5.15

0.442

0.263

1.403

5.00

5.53

0.254

0.317

3.436

4.99

5.81

0.168

0.348

6.740

4.98

Mean

5.00

SIMULTANEOUS SPECTROPHOTOMETRIC DETER-
sect' ■ ^ (CHROMIUM AND MANGANESE). Discussion. This

coiicerned with the simultaneous spectrophotometric determination of
^ solution. The absorbances are additive, provided there is no
ion between the two solutes. We may write:

( 1 )

(2)

763

XVin, 39 QUANTITATIVE INORGANIC ANALYSIS

where and are the measured absorbances at the two wavelengths Aj and Aji
and the subscripts 1 and 2 refer to the two different substances, and the subscripts
Ai and A 2 refer to the different wavelengths. The wavelengths are selected to
coincide with the absorption maxima of the two solutes : the absorption spectra of
the two solutes should not overlap appreciably (compare Fig. XVIII, 40), so that
substance 1 absorbs strongly at wavelength Aj and weakly at wavelength A 2 , and
substance 2 absorbs strongly at A2 and weakly at Aj. Now A = ccl, where e is the
molar absorption coefficient at any particular wavelength, c is the concentration
expressed in mols dm~^ and / is the thickness (length) of the absorbing solution
expressed in cm. If / is 1 cm:

= 2,£i ■ Cl +2,£2 • ^2

^2, =x,£i-Ci+1,£2-C2 (4)

Solution of these simultaneous equations gives:

2, Cl -2/2— 2, £2 -2/ 1

( 5 )

2,Cl-^2, ~2/f^2i

2 .C 1 •2,C2“2,C2-2,Ci

( 6 )

The values of the molar absorption coefficients Ci and Cj can be deduced from
measurements of the absorbances of pure solutions of substances 1 and 2. By
measuring the absorbance of the mixture at wavelengths Ai and Aj, the
concentrations of the two components can be calculated.

The above considerations will be illustrated by the simultaneous
determination of manganese and chromium in steel and other ferro-alloys. The
absorption spectra of O.OOlM-permanganate- and dichromate ions in IM-
sulphuric acid, determined with a spectrophotometer and against IM-sulphuric
acid in the reference cell are shown in Fig. XVIII, 42. The peak at 350 nm for
dichromate solutions cannot be used because iron(lll) ion absorbs strongly
below 425 nm: at a wavelength of 440 nm near the weaker band maximum the
correction for iron(III) ion absorption is small. For permanganate, the
absorption maximum is at 545 nm, and a small correction must be applied for
dichromate absorption. Absorbances for these two ions, individually and in
mixtures, obey Beer’s law provided the concentration of sulphuric acid is at least
0.5M. Iron(III), nickel, cobalt, and vanadium absorb at 425 nm and 545 nm, and
corrections must be made.

Reagents. Potassium dichromate, 0.002M, 0.001 M, and 0.0005M in IM-
sulphuric acid and 0.7M-orthophosphoric acid, prepared from the A.R. reagents.

Potassium permanganate, 0.002M, 0.001 M, and 0.0005M in IM-sulphuric
acid and 0.7M-orthophosphoric acid, prepared from the A.R. reagents. All flasks
must be scrupulously clean.

Procedure, (a) Determination of molar absorption coefficients and
verification of additivity of absorbances.

The molar absorption coefficients must be determined for the particular set of
cells and the spectrophotometer employed. For the present purpose we may
write:

A = ccl

764

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 39

where e is the molar absorption coefficient, c is the concentration (mols dm ^),
and / is the cell thickness or length (cm).

Measure the absorbance A of the above three solutions of potassium
dichromate and of potassium permanganate,, each solution separately, at both
440 nm and 545 nm in 1-cm cells. Calculate e in each case and record the mean
values for CrjO,^" and Mn 04 “ at the two wavelengths.

MixO.OOlM-potassium dichromate and O.OOOSM-potassium permanganate in
the following amounts (plus 1.0 cm^ of concentrated sulphuric acid), and
complete the following table (some typical results are included for guidance only).
Measure the absorbance of each of the mixtures at 440 nm. Calculate the
absorbance of the mixtures from;

^440 —

440eCr-Ccr + 440%n-C|

Test of additivity principle with Cr^O-

and Mn 04 ”

mixtures at

KA 2 O,
Solution, cm^

KMn 04

Solution, cm^

Observed

Calculated

0.371

—

0.338

0.340

0.307

0.308

0.277

0.211

0.214

0.147

0.151

0.086 _

0.088

0.057

—

765

X Vin, 39 QUANTITATIVE INORGANIC ANAL Y SIS

(b) Determination of chromium and manganese in an alloy steel* Weigh out
accurately about I.O g of the alloy steel in a 300-cm^ Kjeldahl flask, add 30 cm^ of
water and 10 cm^ of concentrated sulphuric acid (also 10 cm^ of 85 per cent
phosphoric acid if tungsten is present). Boil gently until decomposition is
complete or the reaction subsides. Then add 5 cm^ of concentrated nitric acid in
several small portions. If much carbonaceous residue persists, add 5 cm^ more of
concentrated nitric acid, and boil down to copious fumes of sulphuric acid. Dilute
to about 100 cm^ and boil until all salts have dissolved. Cool, transfer to a 250-
cm^ graduated flask, and dilute to the mark.

Pipette a 25-cm^ or 50-cm^ aliquot of the clear sample solution into a 250-cm^
conical flask, add 5 cm^ concentrated sulphuric acid, 5 cm^ 85 per cent
phosphoric acid, and 1-2 cm^ ofO.IM-silver nitrate solution, and dilute to about
80 cm^. Add 5 g A.R. potassium persulphate, swirl the contents of the flask until
most of the salt has dissolved, and heat to boiling. Keep at the boiling point for 5-
7 minutes. Cool slightly, and add 0.5 g pure potassium periodate. Again heat to
boiling and maintain at the boiling point for about 5 minutes. Cool, transfer to a
100-cm^ graduated flask, and measure the absorbances at 440 nm and 545 nm in
1-cm cells.

Calculate the percentage of chromium and manganese in the sample. Use
equations (5) and (6) and values of the molar absorption coefficients c determined
above: these will give concentrations expressed in mols dm""’, from which values
the percentages can readily be calculated. Each value will require correction for
the amounts of vanadium, cobalt, nickel, and iron which may be present, using
the following table. The values listed are the equivalent percentages of the
respective constituent to be subtracted from the apparent Cr and Mn percentages
for each 1 per cent of the element in question.

Substance

Cr correction
at 4-10 nm, %

Mn correction
at 545 nm, %

CrjO,^-

0.0025

MnO*'

0.490

—

VO/

0.0266

—

Co^"-

0.0072

0.0011

Ni^*

0.0039

0.0001

Fe^ +

0.0005

—

It can be shown that utilising the known (or determined) molar absorption
coefficients 2.35; 0.01 1 ; 0.369; 0.095):

Mn, percent =
Cr, per cent =

0.00549 F

— ^ (0.426/15^5 - 0.01 3/l,,o)

0.01040F ,

— (2.71/4440—0.1 IO/I 545 )

for a sample of W grams in a volume of V cm^.

* British Chemical Standard No. 225/2 Ni-Cr-Mo steel is suitable for practice in this determination.

766

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 40

Some typical results for Ridsdale’s Alloy Steel, No. 60b, which contained 0.64
per cent Mn, 0.75 per cent Cr, 2.59 per cent Ni, and 0.43 per cent Mo, were;

Percent Mn = 0.63; per cent Cr = 1.10— (0.31+0.01 +0.05) = 0.73.

Experimental.

Determinations by spectrophotometric titrations

XVni,40. SPECTROPHOTOMETRIC TITRATIONS. General discus-
sion. In a spectrophotometric titration the end point is evaluated from data on
. the absorbance of the solution. For monochromatic light passing through a
solution, Beer’s law may be written as :

Absorbance = log I J I, = eel

where Jq is the intensity of the incident light, /, that of the transmitted light, e is the
molar absorption coefficient, c is the concentration of the absorbing species, and I
is the thickness or length of the light path through the absorbing medium. Since
spectrophotometric titrations are carried out in a vessel for which the light path is
constant, the absorbance is proportional to the concentration. Thus in a titration
in which the titrant, the reactant, or a reactive product absorbs radiation, the plot

Fig.XVm,43

Volume of tittant, cm^

of absorbance versus volume of titrant added will consist, if the reaction is
complete and the volume change is small, of two straight lines intersecting at the
end point.

The shape of a photometric titration curve will be dependent upon the optical
properties of the reactant, titrant, and products of the reaction at the wavelength
used. Some typical titration plots are given in Fig. XVIII, 43.

XVIU, 43, A, is characteristic of systems where the substance titrated is
converted into a non-absorbing product.

XVIII, 43, B, is typical of the titration where the titrant alone absorbs.

XVIII, 43, C, corresponds to systems where the substance titrated and the
^*XV*^ colourless and the product alone absorbs.

AVIII, 43, D, is obtained when a coloured reactant is converted into a
CO ourless product by a coloured titrant.

wing to the linear response of absorbance to concentration, an appreciable
rea will often be obtained in a photometric titration, even though the changes
n concentration are insufficient to give a clearly defined inflection point in a
po entiometric titration. Photometric titrations have several advantages over
auh determinations. The presence of other substances absorbing

chan^ wavelength does not necessarily cause interference, since only the
§e in absorbance is significant. The precision of locating the titration line

767

XVm, 401 QUANTITATIVE INORGANIC ANALYSIS

(required for the evaluation of the equivalence point) by pooling the information
derived from several points is greater than the precision of any single point;
furthermore, the procedure may be useful for reactions which tend to be
appreciably incomplete near the equivalence point. An accuracy and precision of
a few tenths per cent are attainable with comparative ease by spectrophotometric
titration. The optimum concentration of the solution to be analysed depends
upon the molar absorption coefficient of the absorbing species involved, and is
usually of the order of 10' molar. The effect of dilution can be made
negligible by the use of a sufficiently concentrated titrant. If relatively large
volumes of titrant are added the effect of dilution may be corrected by multiplying
the observed absorbances by the factor (K+i;)/U, where V is the initial volume
and V is the volume added; if the dilution is of the order of only a few per cent the
lines in the titration plots appear straight. The operating wavelength is selected
on the basis of two considerations: avoidance of interference by other absorbing
substances and need for an absorption coefficient which will cause the change in

absorbance to fall within a convenient
range. The latter is particularly important,
because serious photometric error is poss-
ible in high-absorbance regions. Light
leakage must, of course, be avoided.

The experimental technique is simple.
The cell containing the solution to be
titrated is placed in the light path of a
spectrophotometer, a wavelength appro-
priate to the particular titration is selected,
and the absorption is adjusted to some
convenient value by means of the sen-
sitivity and slit-width controls. A measured
volume of the titrant is added to the stirred
solution, and the absorbance is read again.
This is repeated at several points before the
end point and several more points after the
end point. The latter is found graphically.

XVm, 41. APPARATUS FOR SPECTROPHOTOMETRIC TITRA-
TIONS. A special titration cell is necessary which completely fills the cell
compartment of the spectrophotometer. One for the Unicam SP.500, shown in
Fig. XVIII, 44, can be made from 5-mm Perspex sheet, cemented together with
special Perspex cement, and has dimensions of 9.0 x 9.0 x 4.7 mm. Since Perspex
is opaque to ultraviolet light, two openings are made in the cell to accommodate
circular quartz windows* 23 mm in diameter and 1.5 mm thick: the windows are
inserted in such a way that the beam of monochromatic light passes through their
centres to the photoelectric cell. The Perspex cover of the cell has two small
openings for the tip of a 5-cm^ micro-burette and for a micro-stirrer, respectively
held by means of rubber bungs: the stirrer is ‘sleeved’. The whole of the cell, with
the exception of the quartz windows, is covered with black paper and, as a further
precaution, the top of the cell is covered with a black cloth: it is most important to
exclude all extraneous light.

* These were actually fused silica polarimeter end plates and were supplied by Hilger and Watts Ltd.

768

COLORIMETRY AND SPECTROPHOTOMETRY XVIH, 42/43

XVni,4Z SIMULTANEOUS DETERMINATION OF ARSENIC(m)
and 'ANTIMONY(Iip IN A MIXTURE. Discussion. In acid solution
arsenic(in) can be oxidised to arsenic(V) and antimony(III) to antimony(V) by
the well-established titration with a solution of potassium bromate and
potassium bromide (Section X,. 139). The end-point for such determinations is
usually observed indirectly, and very good results have been obtained by the
spectrophotometric method of Sweetser and Bricker (Ref. 11). No change in
absorbance at 326 nm is obtained until all the arsenic(III) has been oxidised, the
absorbance then decreases to a minimum at the antimony(III) end-point at which
it rises again as excess titrant is added.

Reagents. Bromatefbromide solution. Prepare a standard bromate/
bromide solution by dissolving 2.78 g potassium bromate and 9.9 g potassium
bromide in water and diluting to 1 dm^ in a graduated flask. This solution is
0.017M-potassium bromate (O.lAl) with a slight excess of the theoretical amount
of potassium bromide. Analytical grade reagents should be employed.

Arsenic/antimony solution. Prepare a mixed solution containing approxi-
mately 115 mg arsenic and 160 mg antimonyin 1 dm^ by dissolving about 150 mg
arsenic(III) oxide and 300 mg antimony(III) chloride in 6A/-hydrochloric acid.

Procedure. Place 80 cm^ of the arsenic/antimony solution in the titration
cell of the spectrophotometer. Titrate with standard bromate/bromide solution
at 326 nm taking an absorbance reading at least every 0.2 cm^. From the curve
obtained calculate the concentration of arsenic and antimony in the solution.

XVIII, 43. DETERMINATION OF COPPER(II) WITH EDTA.
Discussion. The titration of a copper-ion solution with EDTA may be carried
out photometrically at a wavelength of 745 nm. At this wavelength the copper-
EDTA complex has a considerably greater molar absorption coefficient than the
copper solution alone. The pH of the solution should be about 2.4.

The effect of different ions upon the titration is similar to that given under
iron(III) (Section XVin, 44). Iron(III) interferes (small amounts may be
precipitated with sodium fluoride solution): tin(IV) should be masked with 20 per
cent aqueous tartaric acid solution. The procedure may be employed for the
determination of copper in brass, bronze, and bell metal without any previous
separations except the removal of insoluble lead sulphate when present.

Reagents. Copper-ion solution, 0.04AT. Wash A.R. copper with A.R.
petroleum ether (b.p.- 40-60°) to remove any surface grease and dry at 100°.

eigh out accurately about 1.25 g of the pure copper, dissolve it in 5 cm^ of
concentrated nitric acid, and dilute to 1 dm^ in a graduated flask. Titrate this
s an ard copper solution with the EDTA solution using Fast Sulphon Black as
EDTA°^ (Section X, 56), and thus obtain a further check on the molarity of the

EDTA solution, O.IOM, and Buffer solution. See Section XVIH, 44.

Procedure. Charge the titration cell (Fig. XVIII, 44) with 10.00 cm^ of the
of fho acetate buffer (pH = 2.2), and about 120 cm^ of
745^'' the spectrophotometer and set the wavelength scale at

StirT 7^?^ width so that the reading on the absorbance scale is zero,
everv 0 so ^^* 3 ^ and titrate with the standard EDTA: record the absorbance
Cont‘ about 0.20 and subsequently every 0.20 cm^.

when *^otil about 1.0 cm^ after the end point; the latter occurs

de absorbance readings become fairly constant. Plot absorbance against

769

xvm,44 QUANTITATIVE INORGANIC ANALYSIS

cm^ of titrant added; the intersection of the two straight lines (see Fig. XVIII, 43,
C) is the end point.

Calculate the concentration of copper ion (mg cm"^) in the solution and
compare this with the true value.

xvni, 44. DETERMINATION OF IRON(in) WITH EDTA. Discussion.
Salicylic aid and iron(III) ions form a deep-coloured complex with a maximum
absorption at about 525 nm: this complex is used as the basis for the photometric
titration of iron(III) ion with standard EDTA solution. At a pH of ca. 2.4 the
EDTA-iron complex is much more stable (higher stability constant) than the
iron-salicylic acid complex. In the titration of an iron-salicylic acid solution with
EDTA the iron-salicylic acid colour will therefore gradually disappear as the
end point is approached. The spectrophotometric end point at 525 nm is very
sharp.

Considerable amounts of zinc, cadmium, tin(IV), manganese(II), chro-
mium(III), and smaller amounts of aluminium cause little or no interference at
pH 2.4: the main interferences are lead(II), bismuth, cobalt(II), nickel, and
copper(II).

Reagents. EDTA solution, O.JOM. See Section X, 50. Standardise
accurately (Section X, 50).

Iron(III) solution, 0.05M. Dissolve about 12.0 g, accurately weighed, of
A.R. ammonium iron(III) sulphate in water to which a little dilute sulphuric acid
is added, and dilute the resulting solution to 500 cm^ in a graduated flask.
Standardise the solution with standard EDTA using Variamine Blue B as
indicator (Section X, 57).

Sodium acetate-acetic acid buffer. Prepare a solution which is 0.2M in
sodium acetate and 0.8M in acetic acid. The pH is 4.0.

Sodium acetate-hydrochloric acid buffer. Add IM-hydrochloric acid to 350
cm^ of IM-sodium acetate until the pH of the mixture is 2.2 (pH meter).

Salicylic acid solution. Prepare a 6 per cent solution of A.R. salicylic acid in
A.R. acetone.

Procedure. Transfer 10.00 cm^ of the iron(III) solution to the titration cell
(Fig. XVIII, 44), add about 10 cm^ of the buffer solution of pH = 4 and about 120
cm^ of water: the pH of the resulting solution should be 1.7-2.3. Insert the
titration cell into the spectrophotometer; immerse the stirrer and the tip of the 5-
cm^ micro-burette (graduated in 0.02 cm^) in the solution. Switch on the tungsten
lamp and allow the spectrophotometer to ‘warm up’ for about 20 minutes. Stir the
solution. Add about 4.0 cm^ of the standard EDTA (note the volume accurately).
Set the wavelength at 525 nm, and adjust the slit width of the instrument so that
the reading on the absorbance scale is 0.2-0.3. Now add 1.0 cm^ of the salicylic
acid solution; the absorbance immediately increases to a very large value (>2).
Continue the stirring. Add the EDTA solution slowly from the micro-burette
until the absorbance approaches 1.8; record the volume of titrant. Introduce the
EDTA solution in 0.05-cm^ aliquots and record the absorbance after each
addition. Continue the titration until at least four readings are taken beyond the
end point (fairly constant absorbance). Plot absorbance against cm^ of titrant
added: the intersection of the two straight lines (see Fig. XVIII, 43, A) gives the
true end point.

Calculate the concentration of iron(III) (mg cm“ ^) in the solution and compare
this with the true value.

770

COLORIMETRY AND SPECTROPHOTOMETRY XVffl, 45/46

Determination of iron(lll) in the presence of aluminium. Iron(III)
(concentration ca. 50 mg per 100 cm^) can be determined in the presence of up to
twice the amount of aluminium by photometric titration with EDTA in the
presence of 5-sulphosalicylic acid (2 per cent aqueous solution) as indicator at pH
1.0 at a wavelength of 510 nm. The pH of a strongly acidic solution may be
adjusted to the desired value with a concentrated solution of sodium acetate:
about 8-10 drops of the indicator solutibii are required. The spectrophotometric
titration curve is of the form shown in Fig. XVIII, 43, A.

xvm, 45. DETERMINATION OF NICKEL ION WITH EDTA.
Discussion. The titration of a nickel-ion solution with EDTA may be performed
photometrically at a pH of about 4.6 and at a wavelength of lOOQ nm, where the
nickel-EDTA complex exhibits characteristic absorption. The titration curve is
similar to that obtained for copper ion (see Fig. XVIII, 43, C). ,

Reagents. Nickel-ion solution, 0.040M. Dissolve about 2.45 g, accurately
weighed, of pure nickel' (Johnson and Matthey) in a mixture of 5 cm^ of
concentrated nitric acid and 5 cm^ of concentrated sulphuric acid, and dilute to 1
dm^ in a graduated flask. Check the concentration of the nickel ion by titration
with standard EDTA solution using Broihopyrogallol Red (Section X, 58) as
indicator. , .

EDTA solution, 0. lOM, and Buffer solutions. See Section XVIII, 44.
Procedure. Proceed as described for Copper (Section XVIH, 43), except
that the buffer solution of pH = 4.0 is employed and the wavelength is adjusted to
1000 nm. The pH of the resulting solution should be about 4.0. Evaluate the end-
point from the titration plots.

Calculate the concentration of the nickel ion (mg cm“^) in the solution and
compare this with the true value.

XVIII, 46, References

1. J. H. Lambert (1760). Photometria sive de Mensura et Gradibus Luminus, Colorum et
Umbrae. Augsburg; reprinted in Ostwald (1892), Kiassiker der Exakten Wissenschaf-
ten. No. 32, 64.

2. M. Bouguer (1729). Essai d’optique sur la Gradation de la Lumiere. Paris; see also
Ostwald (1892), Kiassiker der Exakten Wissenschaften, No. 33, 58. M. Bouguer
(1760). Traite d’optique sur la Gradation de la Lumiere, Ouvrage Posthume. Pub. de
Lacaille.

3. A. Beer (1852). Ann. Pliysik Chem. (J. C. Poggendorff), 86, 78. See also H. G. Pfeiffer
and H. A. Liebhafsky (1951). J. Chem. Educ., 28, 123.

4. F. Bernard (1852). Ann. Chim. Phys., 35, No. 3, 385.

5. D. R. Malinin and J. H. Yoe (1961). J. Chem. Educ., 38, 129.

6. F. H. Lohman (1955). J. Chem. Educ., 32, 155,

7. Determination of Arsenic in Organic Materials (1960). Analytical Methods
Committee, Society for Analytical Chemistry.

8. J. W. McCoy (1969). Chemical Analysis of Industrial Water. MacDonald.

9. H. W. Swank and M. G. Mellon (1938). ‘The Determination of Iron with
Mercaptoacetic Acid’, Ind. Eng. Chem. Anal. Edn., 10.

10 . J. E. Barney,!. G.Bergmann and W.G.Tuskan( 1959). vf«a/.C/ie/M 31 1394

1 1. P. B. Sweetser and C. E. Bricker (1952). ‘Direct SpectrophotometricTitrations with
Bromate-Bromide Solutions’. Anal. Chem., 24, 1107.

771

XVIII, 47 QUANTITATIVE INORGANIC ANALYSIS

XVIII, 47. Selected bibliography

1. E. A. Braude and F. C. Nachod (1955). Determination of Organic Structures by
Physical Methods. Ch. 4. Ultraviolet and Visible Light Absorption. Ch. 5. Infrared
Light Absorption. New York; Academic Press.

2. A. Weissberger(1956). Technique of Organic Chemistry. Vol. 9. Chemical Applications
o/Spectroscopy. New York; Interscience.

3. W. R. Erode and M. E. Corning (1960). Spectrophotometry and Absorptiometry (in
W. G. Berl. Physical Methods of Analysis. Vol. I. 2nd edn.) New York and London;
Academic Press.

4. R. C. Hirt. ‘Ultraviolet Spectrophotometry’, Analytical Chemistry, 1956, 28, 579;
1958,30,589; 1960, 32, 225R.

5. R. F. Godden and D. N. Hume (1954). ‘Photometric Titrations’, Analytical
Chemistry, 26, 1 740.

6. J. B. Headridge(1958). ‘Photometric Titrations’, Talanta, 1, 293.

7. J. B. Headridge (1961). Photometric Titrations. Oxford; Pergamon Press.

8. A. E. Gillam, E. S. Stern and C. J. Timmons (1970). An Introduction to Electronic
Absorption Spectroscopy. 3rd edn. London; Edward Arnold.

9. R. L. Pecsok and L. D. Sheilds (1977). Modern Methods of Chemical Analysis. Ch. 8,
9, 10, 11. New York; Wiley.

10. F. J. Welcher (ed.) (1966). Standard Methods of Chemical Analysis, Vol. IIIB, 6th
edition. New York; Van Nostrand.

11. R. B. Fischer and D. G. Peters (1968). Quantitative Chemical Analysis. Ch. 15 and 16.
Philadelphia; W. B. Saunders.

12. R. C. Denney (1973). A Dictionary of Spectroscopy. London; Macmillan.

13. E, B. Sandell (1959). Colorimetric Determination of Trace Metals. 3rd edn. New York;
and London; Interscience.

772

CHAPTER XIX FLUORIWIETRY*

XIX, 1. GENERAL DISCUSSION. Fluorescence is caused by. the
absorption of radiant energy and the re-emission of some of this energy in the
form of visible light. The light emitted is almost always of higher wavelength than
that absorbed. In true fluorescence the absorption and emission takes place in a
short but measurable time — of the order of 10“^^-10~® second. If the light is
emitted with a time delay (>10“® second) the phenomenon is known as
phosphorescence ; this time delay may range from a fraction of a second to several
weeks, so that the difference between the two phenomena may be regarded as one
of degree only. Both fluorescence and phosphorescence are designated by the
term photoluminescence; the latter is therefore the general term applied to the
process of absorption and re-emission of light energy.

Relationship between intensity of fluorescence and concentration. The
Beer-Lambert law applies to the intensity of radiation transmitted by a
substance or solution ; since fluorescent radiation is that emitted by a substance,
the law cannot be applied directly. The following relationship has been
developed:

F=K{I^-D . (1)

where /„ = intensity of incident radiant energy ;

I = intensity of transmitted radiant energy;

F = intensity of fluorescent radiant energy ; and

A' = a proportionality constant.

F is assumed proportional to the intensity of the radiant energy absorbed (/q — /).

Applying the Beer-Lambert law (Section XVIII, 2),

(2)

and

/o-/=/o(l-10-'0

( 3 )

we have

( 4 )

Writing

( 5 )

F=Fo-Fo.\Q-^'‘

(6)

* The iznafluorometry is usually used in the USA.

773

XIX, 2 QUANTITATIVE INORGANIC ANALYSIS

or log = cic (7)

rQ t

In these equations Kis the fraction of the incident radiation that is absorbed (this
is determined by such factors as the dimensions of the light beam, the area of the
solution irradiated, the transmission band of the filter before the photocell, and
the spectral response of the photocell), e is the molar absorption coefficient and is
dependent upon the substance, / is the thickness of the solution in the cell, and c is
the molar concentration of the fluorescent substance.

When ccl becomes small and approaches a value of 0.01 or less, equation (4)
reduces to:*

F= 2.303 KIo-ccl (8)

or F=K'c (9)

i.e., the fluorescent intensity is practically proportional to the concentration of
the fluorescent substance provided eel ^0.01. cK' is an overall constant for one
particular substance in a given instrument. In practice, equation (8) holds up to a
few parts per million; at higher concentrations the fluorescence-concentration
curve will bend towards the concentration axis.

Factors, such as dissociation, association, or solvation, which would vitiate
the Beer-Lambert law, would be expected to have a similar effect in fluorescence.
Any material that causes the intensity of fluorescence to be less than the expected
value given by equation (8) is known as a quencher, and the e ffe ct is termed
que nching; it is n ormally cau sed bv- the-oresen ce of foreign ion Tof^moleciiles.
FIubre^eiKe_is_affected‘6y tK pH of the solution, bythe nature ofthe'solTent,
the concentration of the reagent which is added in the determination of inorganic
ions, and, in some cases, by temperature. The time taken to reach the maximum
intensity of ffuorescence varies considerably with the reaction.

XIX, 2. INSTRUMENTS FOR FLUORIMETRIC ANALYSIS. Instru-
ments for the measurement of fluorescence are known as fluorimeters or
fluorophotometers. The essential parts of a fluorimeter are shown in Fig. XIX, 1.
The light from a mercury-vapour lamp (a source of ultraviolet light)

Mercury Condensing Primary Sample

vapour lens filter container

* It may be noted that 1 0"“' = e* and that e~^ = +

.'. 1 - 10“' a; fc/fore“' ^ 0.01.

774

FLUORIMETRY XIX, 2

is passed through a condensing lens, a primary filter (to permit the light
band required for excitation to pass), a sample container, a secondary filter
(selected to absorb the primary radiant energy but transmit the fluorescent
radiation), a receiving photocell placed in a position at right angles to the
incident beam (in order that it may not be affected by the primary radiation), and
a sensitive galvanometer or other device for measuring the output of the
photocell. Since fluorescence, intensity is proportional to the intensity of
irradiation, the light source must be very stable if fluctuations in its intensity are
not compensated for. It is usual, therefore, to employ a two-cell instrument ; the
galvanometer is used as a null instrument,' and readings are taken on a
potentiometer used in balancing the photocells against each other. Since the two
photocells are selected so as to be similar in spectral response, it is assumed that
fluctuations in the intensity of the light source are minimised.

The simpler fluorimeters, such as the Locarte,* the EEL 244 f and the Coleman
fluorimeters, are manual instruments operating only at a single selected
wavelength at any one time. Despite this they are perfectly suitable for
quantitative measurements, as these are almost always carried out at a fixed
wavelength. The experiments listed at the end of this chapter have all been carried
out at single fixed wavelengths.

The more advanced fluorescence spectrophotometers are capable of
automatically scanning fluorescent spectra between about 200-900 nm and
produce a chart record of the spectrum obtained. These can also operate at a
fixed wavelength and are equally suitable for carrying out quantitative work,
although their main application tends to be for the detection and determination
of small concentrations of organic substances.

! Locarte Company, 8 Wendell Road, London, W12, England.
T Coming-EEL, Halstead, Essex, England.

775

XEX, 3 QUANTITATIVE INORGANIC ANALYSIS

The optical layout of a typical instrument, the Perkin-Elmer MPF-4, is
illustrated in Fig. XIX, 2. This employs a 1 50 watt xenon lamp power supply and
diffraction grating monochromators. An important difference compared with
spectrophotometers is that in this case the fluorescent radiation is detected by a
photomultipl ier, whereas for absorption spectrpscppy.JbeAefectQiLismsu.aIS[g.
photocell^

Spectrofluorimeters can usually scan at several rates between 10 and
500 nm min” * and give a resolution of the order of 0.5 nm.

XIX, 3. SOME APPLICATIONS OF FLUORIMETRY. Fluorimetry is
generally used if there is no colorimetric method sufficiently sensitive or. selective
for the substance to be determined. In inorganic chemistry the most frequent
applications are for the determination of metal ions as fluorescent organic
complexes, although uranium compounds fluoresce with a brilliant yellow
colour. Uranium may be determined by measuring the fluorescence of a bead
produced by fusing the substance with a mixture of sodium carbonate and
sodium fluoride. Many of the complexes of oxine fluoresce strongly : aluminium,
zinc, magnesium, and gallium are sometimes determined at low concentrations
by this method. Aluminium forms fluorescent complexes with the dyestuff
Eriochrome Blue Black RC (Pontachrome Blue Black R), while beryllium forms a
fluorescent complex with quinizarin.

Important applications are to the determination of quinine and the vitamins
riboflavin (vitamin B^) and thiamine (vitamin B,). Riboflavin fluoresces in
aqueous solution; thiamine must first be oxidised with alkaline hexacyano-
ferrate(III) solution to thiochrome, which gives a blue fluorescence in butanol
solution.

, The intensity and colour of the fluorescence of ma ny substances depend upon
the pH of the solution ; indeed, 'some substances are so sensitive to pH that they
can be use d as pH indicator s. These are termed fluorescent or luminescent
indicators. Those substances which fluoresce in ultraviolet light and change in
colour or have their fluorescence quenched with change in pH can be used as
.fluorescent indicators in acid-base titrations. The merit of such indicators is that
cmTbe” employed' iiTTfie titration of coloured (and sometimes of intensely

I ylJUUA

y they
Table XrX,l

Some fluorescent indicators

Name of indicator

Approx.
pH range

Colour change

Acridine

5.2- 6.6

Green to violet-blue

Chroraotropic acid

3.0- 4.5

Colourless to blue

2- Hydroxycinnamic acid

7.2- 9.0

Colourless to green

3,6-Dihydroxyphthalimide 0.0- 2.5

Colourless to yellowish-green

6.0- 8.0

Yellowish-green to green

Eosin

3.0- 4.0

_ Coloufless-Ucgieen

Erythrosin-B

"XZri' 4.0 ^

Colourless to green

Fluorescein

-4dt- 6.0, Colourless to green

4-Methyl-aesciiretm

”4.0- 6.2 Colourless to blue

9.0-10.0

Blue to light green

2-Naphthoquinoline

4.4- 6.3

Blue to colourless

Quinine sulphate

3.0- 5.0

Blue to violet

9.5-10.0

Violet to colourless

Quininic acid

4.0- 5.0

Yellow to blue

Umbelliferone

6.5- 8.0

Faint blue to bright blue

776

FLUORIMETRY XIX, 4/5

coloured) solutions in which the colour changes of the usual indicators would be
masked. Titrations are best performed in a silica flask. Examples of fluorescent
indicators are given in Table XIX, 1 .

It should be noted that a number of these indicators are also used for other
purposes, e.g., eosin and fluorescein are frequently employed as adsorption
indicators (Section X, 30, C). '

Experimental

XIX, 4. QUININE. Discussion. Although quinine is an organic compound
this determination has been included in this book as it is an ideal experiment.with
which to gain experience in quantitative fluorimetry. It can. be employed
particularly for the determination ofthe amount of quinine in samples of tonic
water.

Reagents. Dilute sulphuric acid, ca. 0.05M. Add 3.0 cm^ concentrated
sulphuric acid to lOOcm^ water, and dilute to 1 dm^ with distilled water.

Standard solution of quinine. Weigh out accurately 0. 100 g quinine and dissolve
it in 1 dm^ 0.05M-sulphuric acid in a graduated flask. Dilute lO.Ocm^ of the
above solution to Idm^ with 0.05M-sulphuric acid.. The resulting solution
contains O.OOlOOmg quinine per cm^. ■

With the aid of a calibrated burette, run 10.0, 17.0, 24.0, 31.0, 38.0, 45.0, 52.0,
and 62.0 cm^ of the above dilute standard solution into separate 100-cm^
graduated flasks and dilute each to the mark with 0.05M-sulphuric acid.

Procedure. Measure the fluorescence of each of the above solutions at
67 1 nm, using that containing 62.0 cm^ of the dilute quinine solution as standard
for the fluorimeter. Use LF2 as the primary filter and gelatine as the secondary
filter if using the Locarte fiuorimeter.

Now prepare test solutioris containing, say, 0.00025 and 0.00045 mg quinine
per cm^. Determine their concentrations by measuring the fluorescence on the
instrument and using the calibration curve.

To determine the quinine content of tonic water it is first necessary to de-gas
the sample either by leaving the bottle open to the atmosphere for a prolonged
period or by stirring it vigorously in a beaker for several minutes. Take 12.5 cm^
of the de-gassed tonic water and make up to 25 cm? in a graduated flask with
O.IM sulphuric acid. From this solution prepare other .dilutions with 0.05M
sulphuric acid until a fluorimeter reading is obtained that falls on the calibration
line previously prepared. From the value obtained calculate the concentration of
quinine in the original tonic water. '

XIX, 5. ALUMINIUM. The procedure utilises Eriochrome Blue Black RC
(also called Pontachrome Blue Black R; Colour Index No. 15705) at a pH of 4.8
in a buffered solution. Beryllium gives no fluorescence and does not interfere;
iron, chromium, copper, nickel, and cobalt mask the fluorescence ; fluoride must
be removed if present. The method may be adapted for the determination of
aluminium in steel.

Reagents. Standard solution, of aluminium. ..Dissolve 1.760 g .A.R.
aluminium potassium sulphate crystals in distilled water, add 3 cm^ concentrated
sulphuric acid, and dilute to 1 dm^ in a graduated flask. Pipette lO.Ocm^ of this
solution into a little water, add 2.0 cm^ concentrated sulphuric acid, and dilute to

777

XIX, 6 QUANTITATIVE INORGANIC ANALYSIS

1 dra^ with distilled water. Tliis solution contains 0.00100 mg aluminium per
dm^.

Ammonium acetate solution, 10 per cent. Dissolve 25 g of the pure salt in water
and dilute to 250 cm^.

Dilute sulphuric acid. Add 25 cm^ concentrated sulphuric acid to 200 cm^
water, cool, and dilute to 500 cm^ in a graduated flask.

Eriocltrome Blue Black RC, 0.1 per cent. Prepare a 0. 1 per cent solution in 90
per cent ethanol.

Procedure. Into 100-cm^ graduated flasks, each containing lOcm^ of the
ammonium acetate solution, 1 cm^ of the dilute sulphuric acid, and 3 cm^ of the
Eriochrome Blue Black RC solution, run in from a burette 15.0, 20.0, 25.0, 30.0,
35.0, 40.0, 45.0, and 50.0 cm^ of the standard aluminium solution. Dilute each of
the above solutions with distilled water, adjust to a pH of 4.6 + 0.2 if necessary
before making to the 100 cm^ mark. Allow the solutions to stand for at least 1
hour.

Measure the fluorescence of each of the above solutions at 590 nm, using that
containing 0.0005 mgcm~^ A1 as standard. The use of a primary filter (Coming
5543 or 5874 has been recommended) will depend upon the quality of the
Eriochrome Blue Black RC; it can often be dispensed with; the secondary filter
may be a Chance OR2 or Corning 2408, or LF7 for the Locarte, Draw a
calibration curve, plotting instrument readings against concentration of
aluminium. Determine the number of mg of A1 per dm^ in an unknown solution
(say, ca. 0.25 mg/dm “^), utilising the calibration curve.

XIX, 6. CADMIUM. Discussion. Cadmium may be precipitated quanti-
tatively in alkaline solution in the presence of tartrate by 2-(2-hydroxyphenyl)-
benzoxazole. The complex dissolves readily in glacial acetic acid, giving a
solution with an orange tint and a bright blue fluorescence in ultraviolet light.
The acetic acid solution is used as a basis for the fluorimetric determination of
cadmium (Ref. 1).

Reagents. 2-(2-Hydroxyphenyl)-benzoxazole solution. Dissolve 1.0 g of
the solid reagent in 1 dm^ of 95 per cent ethanol.

Standard cadmium-ion solution. Prepare a standard cadmium-ion solution
containing ca. 0.04mgcm“^ Cd using A.R. hydrated cadmium sulphate.

Solutions for calibration curve with fluorimeter. Prepare the cadmium complex
of the reagent by precipitating it from a solution of a pure cadmium salt as
follows. Introduce a large excess of sodium tartrate, warm to 60 °C, adjust the pH
to 9-10 by the addition of 0.5A-sodium hydroxide, add a slight excess of the
reagent, and digest at 60 °C for 15 minutes. Filter on a sintered glass crucible
(medium porosity), wash with 50 per cent ethanol (rendered faintly ammoniacal)
to remove excess of the reagent, and dry at 130-1 40 °C for 1-2 hours. Weigh out
0.237 1 g of the complex ( s 0.0500 g Cd) and dissolve it in 1 dm^ of glacial acetic
acid. Remove volumes of the acetic acid solution equivalent to 2.5, 2.0, 1 .0, 0.5,
and O.lOmgCd, and dilute each to exactly 50 cm^ with glacial acetic acid.
Measure the fluorescence of each of the above solutions, using the appropriate
filters (e.g., a yellow ^ter such as Corning 3-74 between the sample and the
photocell). Plot fluorimeter readings against concentration of Cd per 50 cm^.

Procedure. Use an aqueous solution of the sample (25-50 cm^) containing
from 0. 1-2.0 mg of Cd and about 0.1 g of ammonium tartrate. Add an equal
volume of 95 per cent ethanol, warm to 60 °C, treat with a slight excess of the

778

FLUORIMETRY XIX, 7/8

reagent solution (4 cm^ =ca.\ mg Cd), adjust the pH to 9-1 1, digest at 60 °C for
15 minutes, filter on a medium-porosity glass crucible, wash with 20-25 cm^ of 95
per cent ethanol containing a trace of ammonia, and dry the precipitate at 130 °C
for 30-45 minutes. Dissolve the precipitate in 5Q.0 cm^ of glacial acetic acid, and
measure the fluorescence of the solution as in the calibration procedure. Evaluate
the cadmium content from the calibration curve.

XIX, 7. CALCIUM. Discussion. This method is based upon the formation
of a fluorescent chelate between calcium ions and Calcein (fluorescein
iminodiacetic acid) in alkaline solution (Ref. 2). The procedure described below
(Ref. 3) has been employed for the determination of calcium in biological
materials (Ref. 4).*

Reagents. Standard calcium solution. Prepare a standard solution
containing 40.0 mg dm" ^ calcium by dissolving the calculated quantity of
calcium carbonate in the minimum amount of hydrochloric acid and diluting to
1 dm^ in a graduated flask.

Calcein solution. Dissolve sufficient Calcein (fluorescein iminodiacetic acid), or
its disodium salt, in the minimum amount of 0.40M potassium hydroxide
solution and dilute with water to give a concentration of 60 mg dm in a
graduated flask. A small amount of EDTA solution (about 1.0 cm^ of 0.03 A/ for
every 100 cm^ Calcein solution) may be needed in the Calcein solution to achieve
balancing of the blank on the fluorimeter. This is only necessary in those cases in
which the potassium hydroxide used is found to contain a small amount of
calcium impurity.

Aqueous solutions of Calcein are not stable for longer than 24 hours arid
should be kept in the dark as much as possible.

Potassium hydroxide solution: Prepare a 0.4M potassium hydroxide solution
by dissolving solid potassium hydroxide (preferably calcium free) in de-ionised
water and make to 1 dm^ in a graduated flask.

Procedure. Prepare a series of calcium ion solutions covering the
concentration range 0-4 /xg per 25 cm^ by adding sufficient of the 40 mg dm~^
calcium standard to 25 cm^ graduated flasks each containing 5.0 cm^ of 0.4Af
potassium hydroxide solution and icm^ of Calcein. solution. Dilute each to
25 cm^ using de-ionised water. Determine the fluorescence for each solution at
540 nm with excitation at either 330 nm or 480 nm, and plot a calibration curve.

Prepare the sample solution in a similar manner to give a fluorescence value
falling within the range of the calibration curve, and hence obtain the original
calcium concentration in the sample.

XK, 8. ZINC. The zinc complex of oxine fluoresces in ultraviolet light, and
this forms the basis of the following method.

Reagents. Standard zinc solution. Dissolve about 4.0 g, accurately
weighed, A.R. zinc shot in 35 cm^ concentrated hydrochloric acid, and dilute
with distilled water to 1 dm^ in a graduated flask. Pipette 10.0 cm^ of this solution
into a dm^ graduated flask and dilute to the mark with distilled water.

8-Hydroxyquinoline (oxine) solution. 5 per cent. Dissolve 5.0 g A.R. oxine in

f ^ * 0-500 ng range can be determined by using the selective spectrofluorimetric reagent

1 ' J-bis(dicatboxymethyl-aminomethyl)-2,6-dihydroxynaphthalene at pH 1 1 .7 (Ref. 5).

779

XIX, 9/10 QUANTITATIVE INORGANIC ANALYSIS

12 g A.R. glacial acetic acid and dilute to lOOcm^ with distilled water.

' Standard dicldorojiuorescein solution. Add a 0. 1 per cent ethanolic solution of
dichlorofluorescein dropwise to 1 dm^ of distilled water until the resulting
solution has a fluorescence slightly greater than that produced by the most
concentrated zinc solution to be investigated (see below). About 0.8-1. Ocm^ of
dichlorofluorescein solution is required.

Gum arabic solution, 2 per cent. Grind finely 2.0 g gum arable in a glass mortar,
dissolve it in water, and dilute to 100 cm^ ; filter, if necessary.

Ammonium acetate solution, ca. 2M. Dissolve 15.5 g A.R. crystallised
ammonium acetate in water and dilute to 100 cm^.

Procedure. By means of a calibrated burette, run 5.0, 10.0, 15.0, 20.0, and
25.0 cm^ of the standard zinc solution into separate 100-cm^ graduated flasks. To
each flask add 10 cm^ of the ammonium acetate solution, 4cm^ of the gum arabic
solution, dilute to about 45 cm^ with distilled water, and mix by swirling. Now
add exactly 0.40 cm^ of the oxine solution (use, e.g., a micrometer syringe or a
micro-pipette), dilute to the mark with distilled water, shake gently, and transfer
immediately to the cell of a fluorimeter for measurement. Employ the
dichlorofluorescein solution as standard. Use a Chance OB2 as primary filter and
OY2 as the secondary filter. Commence measurements with the most
concentrated zinc solution. It is important that the fluorescence of the zinc-oxine
mixtures be determined immediately after they are prepared, since the fine
suspension of zinc oxinate slowly settles to the bottom of the cell. Plot instrument
readings against zinc content (mg cm"^). Use the calibration curve for
determining the zinc content of test solutions containing, say, 4.5 and 6.5 mg of
zinc perdm^.

XIX, 9. References

1. N. Evcim and L. A. Reber (1954). Anal. Chein., 26, 936.

2. D. F. H. Wallach, D. M. Surgenor, J. Soderberg and E. Delano (1959). Anal. Chem.,
31,456.

3. B. L. Kepner and D. M. Hercules (1963). Anal. Chem., 35, 1238.

4. H. M. von Hattingberg, W. Klaus, H. Lullmann and S. 2^pf (1966). Experientia, 2,
553.

5. B. Budgsmsky and T. S. West (1969). Talania, 16, 399.

XIX, 10. Selected bibliography

1. P. Delahay (1957). ‘Instrumental Analysis’. Ch. 10. Fluorometry, Turbidimetry, and
Nephelometry. New York; The Macmillan Company.

2. H. H. Willard, L. L. Merritt and J. R. Dean (1974). ‘Instrumental Methods of
Analysis’, Ch. 5. Molecular Fluorescence and Phosphorescence Methods. 5th edn. New
York; Van Nostrand Reinhold.

3. J. G. Calvert and J. N. Pitts (1966). Photochemistry. New York; John Wiley.

4. R. B. Cundall and A. Gilbert (1970). Photochemistry. London; Nelson.

5. C. E. White and A. Weissler ( 1 972). ‘Fluorometric Analysis’, Annual Reviews, in Anal.
C/ie/M.,44, No. 5, 182R.

6. J. S. Fritz and G. H. Schenk (1974). Quantitative Analytical Chemistry. Ch. 22, 3rd
edn., p. 438. Boston; Allyn and Bacon.

1. M. Pinta (1974). Detection and Determination of Trace Elements. Chichester; Wiley.

8. G. G. Guilbault (1967). Fluorescence — Theory, Instrumentation and Practice.
London; Edward Arnold, New York; Marcel Dekker.

9. C. E. White and R. J. Argauer (1970). Fluorescence Analysis — A Practical Approach.
New York ; Marcel Dekker.

780

NEPHELOn/IETRY

CHAPTER XX ANDTURBIDIMETRY

XX, 1. GENERAL DISCUSSION. SmalL amounts of some , insoluble
compounds may be prepared in a state of aggregation such that moderately
stable suspensions are obtained. The optical properties of each suspension will
vary with the concentration of the dispersed phase. When light is passed through
the suspension, part of the incident radiant energy is dissipated by absorption,
reflection, and refraction, while the remainder is transmitted. Measurement of
the intensity of the transmitted light as a function of the concentration of the
dispersed phase is the basis of turbidimetric analysis. When the suspension is
viewed at right angles to the direction of the incident light the system appears
opalescent due to the reflection of light from the particles of the suspension
(Tyndall effect). The light is reflected irregularly and diflfusely, and consequently
the term scattered light is used to account for this opalescence or cloudiness. The
measurement of the intensity of the scattered light (at right angles to the direction
of the incident light) as a function of the concentration of the dispersed phase is
the basis of nephelometric analysis (Gr. nephele — a cloud). Nephelometric
analysis is most sensitive for very dilute suspensions (4*100 mg per litre).
Techniques for turbidimetric analysis and nephelometric analysis resemble those
of filter photometry and fluorimetry respectively.

The construction of calibration curves is recommended in nephelometric and
turbidimetric determinations, since the relationship between the ’ optical
properties of the suspension and the concentration of the disperse phase is, at
best, semi-empirical. If the cloudiness or turbidity is to be reproducible, the
utmost care must be taken in its preparation. The precipitate must be very fine, so
as not to settle rapidly. The intensity of the scattered light depends upon the
number and the size of the particles in suspension and, provided that the average
size of particles is fairly reproducible, analytical applications are possible.

The following conditions should be carefully controlled in order to produce
suspensions of reasonably uniform character :

1. The concentrations of the two ions which combine to produce the precipitate
as well as the ratio of the concentrations in the solutions which are mixed.

2. The manner, the order, and the rate of mixing.

3. The amounts of other salts and substances present, especially protective
TOlloids (gelatin, gum arabic, dextrin, etc.).

4. The temperature.

781

XX, 2 QUANTITATIVE INORGANIC ANALYSIS

XX, 2. INSTRUMENTS FOR NEPHELOMETRY AND TURBIDI-
METRY. Visual and photoelectric colorimeters may be used as turbidimeters ;
a blue filter usually results in greater sensitivity. A calibration curve must be
constructed using several standard solutions, since the light transmitted by a
turbid solution does not generally obey the Beer-Lambert law precisely.

‘Visual’ nephelometers (comparator type) have been superseded by the
photoelectric type. It is possible to adapt a good Duboscq colorimeter (Section
XVin, 6) for nephelometric work. Since the instrument is to measure scattered
light, the light path must be so arranged that the light enters the side of the cups at
right angles to the plungers instead of through the bottoms. The usual cups are
therefore replaced by clear glass tubes with opaque bottoms ; the glass plungers
are accurately fitted with opaque sleeves. The light, which enters at right angles to
the cups, must be regulated so that equal illumination is obtained on both sides.
A standard suspension is placed in one cup, and the unknown solution is treated

Fig. XX, 1

in an identical manner and placed in the other cup. The dividing line between the
two fields in the eyepiece must be thin and sharp, and seem to disappear when the
fields are matched.

Most fluorimeters (see Chapter XIX) may be adapted for use in
nephelometry.* The EEL nephelonietert is a simple, inexpensive, and excellent
instrument, and will be described in detail. The essential part of the instrument is
the nephelometer head ; this is illustrated in Figs. XX, 1 , and XX, 2, respectively.
A 6-volt, 6-watt lamp is mounted within the base of this unit to shine light
vertically through the orifice of an annular photocell on to the hemispherical base
of a test-tube. A tri-colour filter wheel (containing filters OB2, ORl, and OGRl)
is interposed between the lamp and the photocell, and is also provided with a
position tor white-light measurement. The selected filter should be similar in

* The Ratio and A4 Fluorimeters manufactured by the Farrand Optical Co. lac, Valhalla, New
York, USA are modified for nephelometry and used in clinical analyses,
t Manufactured by Coming-EEL Ltd, Halstead, Essex, England.

782

NEPHELOMETRY AND TURBIDIMETRY XX, 2

colour to that of the sample. If the solution contained in the test-tube is cloudy of
turbid to any degree the light is scattered by multiple reflections from the
particles. Such scattered light is collected by a reflector, which is mounted above
the photocell, and is then directed on to the photocell itself. The current so
generated is fed by means of a flexible lead and plug to a sensitive galvanometer
(not illustrated). A metal cap is provided to fit. over the test-tube when the
instrument is in use and will exclude extraneous light. When this cap is removed a
micro-switch is operated to disconnect the photocell from the galvanometer and
prevent damage to the suspension by large current resulting from the sudden
entry of external light. The sensitive galvanometer is of the taut-suspension
mirror type: a stabilising transformer is incorporated to supply power to the
nephelometer lamp. A large plastic knob protrudes through the top of the

galvanometer casing to provide a smooth zero setting ; a sensitivity control and a
clamping device are also incorporated. The standard size of matched test-tube is
1.5 cm diameter and 15 cm long, but adapters are available for test-tubes of
1 .25 cm and 1 .9 cm diameter. A Perspex standard is supplied, and can be used for
standardising the nephelometer; this standard may be immersed in liquids of
various refractive indices, if desired.

The general procedure for operating the EEL nephelometer is as follows :

1. Adjust the zero control knob of the galvanometer to bring the hair-line to
the zero of the scale. Connect the nephelometer head and galvanometer unit by
means of the six-pin plug and flexible lead.

2. Remove the cap cover, place the standard in position, and replace the cap.

783

XX, 3 QUANTITATIVE INORGANIC ANALYSIS

The standard is generally the matched test-tube containing the most
concentrated suspension of the substance being determined : the concentration
must, of course, be known. ■

3. Select the filter required. This should be similar in colour to that of the

solution. -

4. Adjust the sensitivity control of the galvanometer to obtain a reading of one
hundred divisions on the scale.

5. Remove the standard.

6. Fill a test-tube with distilled water or with a ‘blank.’ solution to a depth of
not less than 3 cm, and set to zero by means of the galvanometer zero control.

7. Check the reading ofthe standard (100 divisions).

8. Repeat procedures 6 and 7 until full-scale deflection and zero settings are
obtained.

9. Replace the standard suspension with more dilute suspensions, and note
the various scale readings. Draw a calibration curve relating galvanometer
readings and the concentrations of the substance being determined.

10. Fill a test-tube with the sample to be determined to a depth similar to that
used for the standards, and insert into the instrument. Note the galvanometer
deflection', evaluate the concentration from the calibration curve.

Some nephelometric determinations

XX, 3. SULPHATE. Discussion. The turbidity of a dilute barium sulphate
suspension is difficult to reproduce; it is therefore essential to adhere rigidly to
the experimental procedure detailed below. The velocity of the precipitation, as
well as the concentration of the reactants, must be controlled by adding (after all
the other components are present) pure solid barium chloride of definite grain
size. The rate of solution of the barium chloride controls the velocity of the
reaction. Sodium chloride and hydrochloric acid are added before the
precipitation in order to inh ibit the growth of microcrvstals of barium .sji lphate;
the optimum pH is maintaine^and minimises the effect of variable amounts of
other electrolytes present in the sample upon the size of the suspended barium
sulphate particles. A glycerol-ethanol solution helps to stabilise the turbidity.
The reaction vessel is shaken gently in order to obtain a uniform particle size:
each vessel should be shaken at the same rate and the same number of times. The
unknown must be treated exactly like the standard solution. The interval
between the time of precipitation and measurement must be kept constant.

Reagents. Standard sulphate solution. Dissolve 1.814 g dry A.R.
potassium sulphate in distilled water and dilute to 1 dm^ in a graduated flask.
This solution contains 1.000 mg of sulphate ion per cm^.

Sodium chloride-hydrochloric acid reagent. Dissolve 60 g A.R. sodium chloride
in 200 cm^ distilled water, add 5cm^ pure concentrated hydrochloric acid, and
dilute to250cm^.

Barium chloride. Use crystals of A.R. barium chloride that pass through a 20-
mesh sieve and are retained by a 30-mesh sieve.

Glycerol-ethanol solution. Dissolve 1 volume of pure glycerol in 2 volumes of
absolute ethanol.

Procedure. Run 0.5, 1.0, 1.5, 2.0, 2.5, and 3.0 cm^ of the standard
potassium sulphate solution from a calibrated burette into separate 100-cm^

784

NEPHELOMETRY AND TURBIDIMETRY XX, 4

graduated flasks. To each flask add 10 cm^ of the sodium chloride-hydrochloric
acid reagent and 20 cm^ of the glycerol-ethanol solution, and dilute to lOOcm^
with distilled water. Add 0.3 g of the sieved barium chloride to each flask, stopper
each flask, and shake for 1 minute by inverting eaeh flask once per second : all the
barium chloride should dissolve. Allow each flask to stand for 2-3 minutes and
measure the turbidity in the EEL nephelometer: take care to avoid small air
bubbles adhering to the walls of the matched test-tubes. Use the most con- ■
centrated solution as standard and, by means of the sensitivity control, adjust
the galvanometer, reading to 100 divisions. Prepare a ‘blank’ solution; repeat the
above sequence, of operations, but do not add any sulphate solution. Place
the ‘blank’ solution in the nephelometer and- adjust to zero reading of the
galvanometer scale by means of the zero control above the galvanometer
suspension. Check the reading of the most turbid solution, and adjust any
deviation from 100 by means of the sensitivity control. Repeat the measurements
with the flve other standard sulphate solutions. Plot the galvanometer reading
against the sulphate-ion content per cm^.

Determine the sulphate-ion content of an unknown solution, say, ca.
0.5 mg cm“ ^ : use the calibration curve. .

XX, 4. PHOSPHATE. Discussion. Phosphate ion is determined nephelo-
metrically following the formation of strychnine molybdophosphate. This
turbidity is white in colour and consists of extremely fine particles (compare
ammonium molybdophosphate, which is yellow and is composed of rather large
grains). The precipitate must not be agitated, as it tends to agglomerate easily; it
is somewhat sensitive to temperature changes.

Reagents. Standard phosphate solution. Dissolve 1.721 g A.R. potassium
dihydrogenphosphate (dried at 1 10 ‘’C) in 1 dm^ of water in a graduated flask.
Pipette lO.Ocm^ of this solution into a 1-dm^ graduated flask and dilute to the
mark. The resulting dilute solution contains 0.01 mg phosphorus pentoxide per
cm^.

Molybdate-strychnine reagent.* This reagent is prepared in two parts; these
are mixed just before use, since the addition of the acid molybdate solution to the
strychnine sulphate solution produces a precipitate after 24 hours. Solution A
(acid molybdate solution): place 30 g A.R. molybdenum trioxide in a 500 cm^
conical flask, add 10 g A.R. sodium carbonate and 200 cm^ water. Boil the
mixture until a clear solution is obtained. Filter the hot solution, if necessary.
Add 200 cm^ 5M-sulphuric acid, allow to cool, and dilute to 500 cm^.

Solution B (strychnine sulphate solution) : dissolve 1 .6 g strychnine sulphate
in 100 cm^ warm distilled water, cool and dilute to 500 cm^.

Prepare the reagent by adding Solution B rapidly to an equal volume of
Solution A, and shake the resulting mixture thoroughly; filter off the bluish-
white precipitate through a Whatman No. 42 filter paper. The resulting clear
solution will keep for about 20 hours. Solutions A and B may be kept indefinitely,
sn sulphate solution. Prepare a saturated aqueous solution at

0 C and cool to room temperature. Filter before use.

^ alkaloid. It should only be handled with gloves and under no circumstances
Should It be ingested.

785

XX, 5 QUANTITATIVE INORGANIC ANALYSIS

Sulphuric acid, IM. Dilute 27 cm^ of A.R. concentrated sulphuric acid to
500 cm^ in a graduated flask.

Procedure. Run in 1.0, 2.0, 4.0, 6.0, 8.0, and lO.Ocm^ of the standard
phosphate solution from a calibrated burette into separate 100-cm^ graduated
flasks. To each flask add 18cm^ IM-sulphuricacid and I6cm^ saturated sodium
sulphate solution, and dilute to approximately 95 cm^ with distilled water. Now
add 2.0 cm^ of the molybdate-strychnine reagent and dilute to 100 cm^ ; mix the
contents of the flask by gently inverting several times, but do not shake. Allow the
flasks to stand for 20 minutes to permit the turbidities to develop before making
the measurements. Prepare a ‘blank’ solution by repeating the above sequence of
operations, but omit the addition of the phosphate solution. Use the most
concentrated solution as the initial standard and adjust the galvanometer reading
to 100 divisions. Introduce the ‘blank’ solution into the matched test-tube of the
EEL nephelometer and adjust the galvanometer reading to zero. Check the
standard solution for a galvanometer reading of 100. Repeat the above with
the five other phosphate solutions. Plot galvanometer reading against mg PjOj
per cm^.

Determine the phosphate content of an unknown solution, say, containing ca.

0.005 mg PjOj per cm^ ; use the calibration graph.

XX, 5. Selected bibliography

1. R. Barnes and C. R. Stock (1949). ‘Apparatus for Transmission Turbidimetry of
Slightly Hary Materials’, Analytical Chemistry, 21, 18.

2. C. L. Wilson (1953). ‘Nephelometry’, Annual Reports on the Progress of Chemistry, 50,
367.

3. P. Delahay (1957). Instrumental Analysis. Section on Turbidimetry and Nephelom-
etry. New York; The Macmillan Company.

4. G. W. Ewing (I960). Instrumental Methods of Chemical Analysis. Section on
Nephelometry and Turbidimetry. 2nd edn. New York; McGraw-Hill Book Co.

5. H. A. Strobel (1960). Chemical Instrumentation. A Systematic Approach to
Instrumental Analysis. Ch. 8. Light Scattering Photometry. Reading, Mass. ; Addison-
Wesley Publishing Co.

786

CHAPTER XXI EIMISSIOM SPECTROGRAPHY

XXI, 1. GENERAL DISCUSSION. When certain metals are introduced as
salts into the Bunsen flame characteristic colours are produced ; this procedure
has long been used for detecting elements qualitatively. If the light from such a
flame is passed through a spectroscope several lines may be seen, each of which
has a characteristic colour; thus calcium gives red, green, and blue radiations, of
which the red are largely responsible for the typical colour that this element
imparts to the flame. A definite wavelength can be assigned to each radiation,
corresponding with its fixed position in the spectrum. Although the flame colours
of, for example, calcium, strontium, and lithium are very similar, it is possible to
differentiate with certainty between them by observations on their spectra and to
detect each in the presence of the others. By extending and amplifying the
principles inherent in the qualitative flame test, analytical applications of
emission spectrography have been developed. Thus more powerfiU methods of
excitation, such as electric spark or electric arc, are used, and the spectra are
recorded photographically by means of a spectrograph: also, since the
characteristic spectra of many elements occur in the ultraviolet, the optical
system used to disperse the radiation is generally made of quartz.

A detailed discussion of the origin of emission spectra is beyond the scope of
this book but a simplified treatment is given in Chapter XXII, Sections 1 and 2 *

It may be stated, however, that there are three kinds of emission spectra :
continuous spectra, band spectra, and line spectra. The continuous spectra are
emitted by incandescent solids, and sharply defined lines are absent. The band
spectra consist of groups of lines that come closer and closer together as they
approach a limit, the head of the band : these are caused by excited molecules.
Line spectra consist of definite, usually widely and seemingly irregularly spaced,
lines ; these are characteristic of atoms or atomic ions which have been excited
and emit jhmfener^diLtheJom of light of definite wavelengths. The quantum
theory predicts that each atonTbr ion possesses definite energy states in which the
vanous electrons can exist; in the normal or ground state the electrons have the
lowest energy. Upon the application of sufficient energy by electrical, thermal, or
other means, one or more electrons may be removed to a higher energy state
arther from the nucleus ; these excited electrons tend to return to the ground

For a more detailed treatment of the theory of atomic spectroscopy the reader is referred to G.
erz erg (1944). Atomic Spectra and Atomic Structure. 2nd edn. Dover Publications.

787

XXI, 1 QUANTITATIVE INORGANIC ANALYSIS

state, and in so doing emit the extra energy as a photon of radiant energy. Since
there are definite energy states and since only certain changes are possible
according to the quantum theory, there are a limited number of wavelengths
possible in the emission spectrum. The greater the energy of the exciting sourcCj^
the higher the energy of the excited electrons, and therefore the more numerous
the lines that may appear. The intensity of a spectral line depends largely upon
the probability of the required energy transition or ‘jump’ taking place. The
intensity of some of the stronger lines may occasionally be decreased by |elf>^
absorption caused by reabsorption of energy by the cool gaseous atoms in the-
outer regions of the source. With high-energy sources the atoms may be ionised
by the loss of one or more electrons; the spectrum of an ionised atom is different
from that of a neutral atom and, indeed, the spectrum of a singly ionised atom
resembles that of the neutral atom with an atomic number one less than its own.

The lines in the spectrum from any element always occur in the same positions
relative to each other. When sufficient amounts of several elements are presentin
the source of radiation each emits its characteristic spectrum ; this is the basis for
qualitative analysis by the spectrochemical method. It is not necessary to
exammHlnd'identify all the lines in the spectrum, because the strongest lines will
be present in definite positions, and they serve to identify unequivocally the
presence of the corresponding element. As the quantity of the element in the
source is reduced, these lines are the last to disappear from the spectrum: they
have therefore been called the persistent lines or the ‘raies ultimes’ (R.U. lines),
and simplify greatly the qualitative examination of spectra.

Lines in an unknown spectrum may be identified by comparing them with
those on a spectrum containing a number of lines of known wavelengths. This
may be performed either by comparison with charts of spectra of metallic
elements such as iron or copper, or by the use of R.U. powder (see Section
XXI, 3).

The number of lines appearing in the spectrum varies considerably from
element to element. The spectra of the transition elements, the lanthanoids and
such elements as titanium and molybdenum, produce complex spectra; copper,
antimony, tin, and lead are intermediate; while boron, magnesium, aluminium,
zinc, and the alkaline-earth metals give relatively simple spectra. The practical
result of these differences is that spectrographs with greater dispersion and
resolution are required to separate adequately the lines in complex spectra, e.g.,
iron, nickel, cobalt, or manganese: such spectrographs are necessarily large and
expensive.

I For quantitative analysis it is necessary to assess the densities of blackening of
llines in a spectrogram due to the constituents being determined ; this may be done
by comparing the spectra from samples of known and unknown composition.
Comparisons may be made either visually (best with the aid of a spectrum
projector: see Fig. XXI, 6) when no great accuracy is desired, or by photoelectric
measurement of line densities with a microphotometer (see Section XXI, 2).
Details of the procedure are described in Sections XXI, 4, and XXI, 7.

The applications of emission spectrography include :

1 . the examination of a single metal or an alloy for impurities ;

2. the analysis of an alloy for its general composition, including a search for

minor components and traces of impurities ;

3. the analysis of ash of organic substances and other materials (e.g., natural

waters) amenable to similar treatment ; and

788

EMISSION SPECTROGRAPHY XXI, 2

4. the detection of contaminants in food.

The chief advantages of the spectrographic method of analysis are :

(a) The procedure is specific for the element being determined, although
difliculties occasionally arise when a line of another element overlaps that of
the unknown.

(b) The method is time-saving; a quantitative determination of traces of the
elements in a sample, especially an alloy or a metal, may be made without any
preliminary treatment. Most metals and some non-metals (e.g., phosphorus,
silicon, arsenic, and boron) may be determined.

It may be (and is usually) applied to the determination of small quantities of
added constituents or of traces of impurities where conventional methods of
analysis are difficult, fail, or give less accurate results. Lengthy and difficult
separations by chemical methods, e.g., of zirconium and hafnium and of
niobium and tantalum, can be avoided.

The apparent disadvantages are :

(i) Successful use requires wide experience, both in the operation of equipment
and in reading and interpreting spectra.

(ii) The spectrograph is essentially a comparator; for quantitative analysis,
standards (usually of similar composition to the material under analysis) are
necessary. Unknown samples therefore present a relatively difficult problem
when quantitative results are required.

(iii) The accuracy and precision are not as high as gravimetric, titrimetric, and
some spectrophotometric methods for elements present in quantities greater
than 2-5 per cent of the total ; indeed, spectrographic methods are not usually
applied for elements present to a greater extent than about 3 per cent.

XXI, 2. EQUIPMENT FOR EMISSION SPECTROGRAPHIC ANALYSIS.
This section is concerned with describing the equipment which is necessary
for an introduction to spectrographic techniques for the analyst. In this instance
the practical work will be described for instruments manufactured by Rank
Hilger, Margate, Kent, England, but the comparable products of other
manufacturers (e.g., Bausch and Lomb, Rochester, USA) may also be used.

The essential parts of a spectrograplLUFe-a slit, an optical system, and a
camera forreCOrdingThe specirumTThe light from the source of radiation passes
through the slit, which is a narrow vertical aperture, then through the optical
system, which includes a prism or grating. An image of the slit is produced by
means of lenses at the point where the light is recorded. One such image is
produced for each radiation having a specific wavelength, and the result is a series
of vertical line images which constitute the spectrum of the element being
investigated. The optical system may be either of glass or quartz; the latter
ransmits in the ultraviolet region, where many useful lines occur, as well as in the
^10000^*^°^^ of wavelengths employed with quartz extends from about 2000

Two sizes of prism spectrographs are widely used for analysis, the medium and

should require that these values are expressed in nanometres;
particular employed and it is felt desirable to retain it for this

789

XXI, 2 QUANTITATIVE INORGANIC ANALYSIS

Fig. XXI, 1

the large. The large spectrograph is necessary for the analysis of iron, chromium,
cobalt, molybdenum, titanium, tungsten, uranium, and zirconium owing to the
numerous lines in the spectra and the need for maximum resolution; it utilises a
25.4 cm X 10.2 cm (10 in. x 4 in.) plate, but adjustments must be made to bring
different regions of the spectrum on the plate. The complete spectrum is 76 cm
long, so that it is recorded in three separate sections. The Hilger and Watts
medium spectrograph has ample dispersion for most work with non-ferrous
metals and light alloys : it has the advantage that the whole of the spectrum range
can be photographed in a single exposure on one plate (25.4 cm x 10.2 cm). The
essential features are shown in Fig. XXI, 1, while the actual instrument is
depicted in Fig. XXI, 2. The light produced from the sample by either means of
excitation enumerated below is received by a narrow slit and passes through a
lens system to a prism which deviates each radiation from a direct path by an
amount depending upon its wavelength. A second lens system forms an image of
the narrow slit upon a photographic plate in the order determined by the prism.
The prism employed is known as a Cornu prism;* the 60° prism is composed of
two half prisms of quartz in optical contact. The two halves are cut so that they
compensate each other, with the result that the combination functions as if it
were an equilateral quartz prism.

The slit (20 mm long) is fonned of two parallel metal jaws, which are accurately
ground flat ; the width is controlled to 0.001 mm by means of a micrometer screw
which causes both jaws to move symmetrically. For most determinations with
spectrographs a slit width of about 0.02 mm is usually satisfactory. The slit is
equipped with a Hartmann diaphragm (see Fig. XXI, 15) for placing spectra in
juxtaposition on the plate; this involves a number (usually three) of square
openings arranged in echelon, the square holes being cut with the bottom of one
in line with the top of the next. When these holes are placed successively in front
of the slit it is possible to record three spectrograms in juxtaposition without
moving the plate holder. The shutter is usually placed between the slit jaws and is
operated by a lever.

* The Littrow prism is also widely used. It is a 30° prism with a mirror back face; the light passes
through the prism to the mirror face and is reflected back through the prism, the total path being
equal to a 60° prism. The prism mounting results in a fairly compact instrument.

790

EMISSION SPECTROGRAPHY XXI, 2

Fig. XXI, 2

The spectrograph is provided with a scale graduated in wavelengths, which can
be illuminated and printed directly on the spectrogram plate. It is incorporated in
the back of the camera and is controlled by a lever. The camera slide for carrying
the plate holder is fitted at the end of the instrument casting. Provision is made
for turning the plate-holder carrier mounting through a small angle about a
vertical axis. The plate-holder slide is operated by a rack-and-pinion motion that
raises or lowers the plate carrier over a range of 75 mm in 1-mm steps. A number
of spectra (up to 40, but depending on the length of the slit) can be taken on one
plate.

The length of the instrument from the slit to the end of the plate holder is about
1.2 metres, and it is supported on a massive base which raises the optical parts
about 30 cm above bench level. An optical bar of steel is attached to the base of
the instrument, from which it projects about 90 cm ; it is parallel with the optical
axis. The bar serves to carry lenses, an arc and spark stand (Gramont stand) for
holding samples, and other ancillary equipment.

The holders for arc and spark excitation fit directly on the optical bar attached
to the spectrograph. In modem instruments the electrode holders are housed in a
metal box fitted with a safety shield. The electrodes are carried in strong screw
clamps on horizontal arms which are insulated from the supporting base ; the
movements include vertical adjustment of the upper electrode with rack and
pinion, vertical and horizontal movements of the lower electrode through a
collar which may be clamped to the supporting rod, and vertical and rotary
movements of both electrodes together. The combinations are such that the
discharge can be rapidly located on the optical axis. The holders are primarily
designed for cylindrical rods not greater than 1 .25 cm diameter.

The slit should be uniformly illuminated along its length. For this purpose a
quartz 1ms, of such focal length that it throws an image of the source on the
collimating lens (see Fig. XXI, 1) is placed 2 cm from the entrance slit and located

I ® 1

b+ [1

Arc gap ^

0 -

vyv^vv 'ItCWgtRT'

%-XXI,3

791

XXI, 2 QUANTITATIVE INORGANIC ANALYSIS

by means of the optical bar. The rapid location of the source of light (i.e.,
positioning of the electrodes) on the optical axis of the spectrograph and the
adjustment of the length of the discharge gap is conveniently carried out by
projecting an image of the source on to a calibrated screen (gauge plate) provided
with the instrument. Both the lens and the screen are mounted on the optical bar
at the end remote from the slit (Fig. XXI, 12). After the initial adjustment with the
light source, a reading lamp is arranged so that the lampi can be brought quickly
into position near the electrodes on the slit side of the stand; it will be found that
the electrodes can be seen on the screen. For subsequent work the electrodes can
be set in optical alignment with the spectrograph simply by setting the image of
the electrodes to agree with the screen gauge without actually passing any current
between them. The two outer horizontal lines on the screen gauge are equivalent
to an electrode gap of 4 mm and the inner lines are equivalent to a 2-mm gap.

The two most commonly used excitation sources are the low-voltage d.c. arc
and the high-voltage a.c. spark. For the low-voltage d.c. arc the essential
requirements are a source of d.c. at 1 10-250 volts, a regulating resistance R to
control the current (2-12 amperes), an inductance L (this tends to steady the arc
and maintain a more constant voltage), a d.c. ammeter A in series with, the
supply, and an arc gap (see Fig. XXI, 3). The sample under investigation may
consist of electrically conducting rods, which then become the electrodes of the
arc. In order to strike the arc, the two electrodes must be brought into contact or
can be short-circuited by touching them with an insulated third (carbon)
electrode. If the sample is in the form of a powder or small pieces it may be
supported on a pure carbon or graphite rod which has been hollowed out to hold
the specimen. Another similar rod, generally with a pointed or conical tip, serves
as the upper electrode. Sometimes pure metal rods, e.g., copper or silver, are used
as the supporting electrodes, and the sample may be made either the cathode or
the anode. It is advisable to make exposures with the sample respectively positive
or negative in order to determine which gives the better results for reproducibility
and sensitivity. With two carbon electrodes it is usual to place the sample in the
anode (positive electrode), which serves as the lower electrode. The temperature
in the arc stream ranges from 2000 °C to about 5000 °C. When an arc is operated
between carbon electrodes in air, some cyanogen molecules are formed, and these
may emit molecular band spectra in the region 3200-4200 A.

The d.c. arc is a sensitive source and of wide application, but its reproducibility

is not of the highest order; it is generally used for the identification and
determination of elements present in very small concentrations. A comparatively
large amount of the substance being analysed passes through the , arc, and

792

XXI, 2 QUANTITATIVE INORGANIC ANALYSIS

focusing lenses enable clear definitions of the individual spectra to be obtained
and a ten-fold magnification is possible. Accurate alignment of the spectra is
possible by means of a fine adjustment screw.

For quantitative analysis it is necessary to compare the relative blackening of
lines with one another and with those produced by standard elements. The
density of blackening (or simply blackening) B may be defined as :

B = log^

where /q is the intensity of the transmitted radiation by a perfectly clear part of
the photographic plate and / is the light transmitted by the line in question.

An apparatus for measuring the densities of blackening of very small areas of
photographic plates or films is called a microphotometer. The essential features
of this type of instrument are shown in Fig. XXI, 7, and an actual instrument, the
Rank Hilger L500 Microphotometer, is illustrated in Fig. XXI, 8.

Plate

Full instructions for use are supplied with the instrument but the main features
are as follows. A reduced image of the filament of the lamp is focused on the
emulsion side of the photographic plate. An enlargement (10 x ) of the plate is
projected on the whitened jaws of the slit. The light passing through the slit is
detected by a photocell whose output can be shown either as a galvanometer

794

EMISSION SPECTROGRAPHY XXI, 3

Fig. XXI, 8

deflection or by a photomultiplier from which the output is fed to a chart
recorder. The slit width of the microphotometer is controlled by a drum
graduated to read to 0.005 mm; the slit height can be reduced by a milled-edged
control situated in front of the micrometer drum. An additional lens can be
moved into the light path by ineans of a lever in front of the instrument. The effect
of this lens is to flood a small area of the plate with light in order that an image of
a short length of the spectrum is seen on the slit-jaw screen and the line to be
photometered can be adjusted into the correct position. When the lens is removed
to take a reading, a shutter opens before the barrier layer cell or photomultiplier,
which is then exposed to light only when the actual readings are taken. The base
of the microphotometer supports guide rods on which the plate stage travels.
This stage accommodates plates up to 25.4 cm in length and 10.2 cm wide. To
bring the spectrum into position the plate holder can be moved at right angles to
the traverse by a rack and pinion gearing. A motor-driven leadscrew can move
the plate stage to the left or right at a traversing speed of 0. 1 mm min ~ ^ , but this
may be overriden by means, of a manual control, even during a motor-driven
traverse.

^1,3. qualitative SPECTROGRAPHIC ANALYSIS. At least fifty-
ve elements can be identified under normal conditions of excitation. In a
qualitative analysis it is desirable to have a high sensitivity so that the presence of
race elements may be revealed. The d.c. arc usually gives the highest sensitivity,
n many cases satisfactory results may be obtained with the a.c. high-voltage
park as the source of excitation; although the sensitivity is lower, the
excT W greater. The d.c. arc is preferred for the most difficultly

bura^ f non-volatile and refractory compounds ; a ‘complete

01 the sample may be obtained with comparative ease. A widely used

795

XXI, 3 QUANTITATIVE INORGANIC ANALYSIS

system of qualitative spectrographic analysis is to arc the sample in a depression
in a graphite electrode, using a pointed counter electrode. The graphite electrode
in air gives a cyanogen band spectrum (headings at 4216, 3883, and 3590 A) in the
near ultraviolet, which may mask some of the lines of interest. A sample of a few
milligrams is sufficient; the exposure conditions must be determined by
experiment. Non-conducting samples may be more easily excited by being mixed
with an equal volume of graphite powder, and these, as well as all refractory
materials, are often treated with a carrier to bring them into the arc stream. This
carrier, usually a volatile salt such as ammonium chloride or ammonium
sulphate, helps to propel the entire sample smoothly up into the arc gap.

The amount of an element that is detectable varies with its concentration, its
relative volatility, the energy of excitation, etc. : approximate sensitivity figures
with arc excitation for the common elements are collected in Table XXI, 1 .

As one dilutes the amount of an element in an arc, the number of lines
observable is reduced, and ultimately there remains only a few lines of the
element which is diluted. These lines are known as the ‘rales ultimes’ or persistent
lines. Tables of these persistent lines may be found in the references in the
Selected Bibliography at the end of this chapter, in chemical handbooks, and in
Appendix XIV. The identification of these lines will permit detection of elements
present in low concentration, and all qualitative methods utilise the persistent
lines.

Table XXI, I Arc sensitivities of some elements

O.l-I per cent delectable As, Cs, Nb, P, Ta, W.

0.01-0.1 per cent detectable B, Bi, Cd, La, Rb, Sb, Si, Tl, Y, Zn, 2r. ■ '

0.001-0.01 per cent detectable Al, Au, Ba, Be, Ca, Fe, Ga, Ge, Hg, Ir, K, Mn, Mo, Pb, Sc, Sn,

Sr,Ti,V.

0.0001-0.001 per cent detectable Ag, Co, Cr, Cu, In, Li, Mg, Na, Ni, Os, Pd, Pt, Rh, Ru.

The simplest and most direct procedure for the qualitative analysis of an
unknown sample is the R.U, powder method. The R.U. powder is a powder
developed in the Research Laboratories of the General Electric Company of
Wembley, England, and is marketed by Johnson, Matthey and Co of London. It
consists of small quantities of fifty elements incorporated in a base material
composed of calcium, magnesium, and zinc oxides. The quantity of each element
present has been carefully adjusted so that only the ‘raies ultimes’ and the most
important sensitive lines appear when the spectrum is excited by placing some of
the powder ( 1 0-20 mg) in the lower (and positive) pole of an arc between graphite
electrodes. A current of 5-7 amperes and an arc length of 6 mm is recommended.
It is advisable not to expose for a longer time than the powder lasts, or else parts
of the spectrum will be unnecessarily masked by the CN bands. A set of seven
enlargements of the arc spectra of the R.U. powder is marketed, and these cover a
wavelength range of 2284-8000 A. The spectrum of the iron arc (together with
the important wavelengths) is given alongside that of the R.U. powder; this
enables the positions of the persistent lines relative to the iron-arc spectra to be
seen immediately, and will also permit the position of any sensitive line to be
found readily. A qualitative analysis is carried out by producing contiguous arc
spectra with the aid of a Hartmann diaphragm of the R.U. powder,- the sample,
and optionally (if known) of a pure main element present in the sample. The plate
is developed, and the lines present in both the R.U. powder and sample spectra

796

EMISSION SPECTROGRAPHY XXI, 3

are noted; the latter is most, simply observed on a Rank Hilger projection
comparator (Fig. XXI, 6). The elements are tabulated with the number of lines
appearing: three lines,* which are free from interference, are considered proof of
the presence of the element. A portion of the spectrum of R.U. powder and a
sample of Wood’s metal is given in Fig. XXI, 9; both major and minor
constituents are readily identified .

Another problem that frequently arises is to decide whether a substance
contains a given element or a small number of specified elements. As an example
we may take the presence of cadmium in spelter. Contiguous spectra are taken
using the Hartmann diaphragm of : (i) the spectroscopically pure metal known to
be present in the specimen under test in considerable quantity (e.g., zinc) ; (ii) the
sample under test (e.g., spelter) ; and (iii) the spectroscopically pure metal whose

SnZnlln

Pi Bi Zn A1 Mg V

Bi BiCd

Sn V

Ca Mo 1 Ca j Mo Ti Mo Ca
I S |lnU I r* H S

Ti CuInCdlCuAgZn
rtn I (nil — * r* A — I

Cd Sn

Sn V

R.U. Powder
Woods metal

SnPb CuInCdSnCu Ag

SOUA 3282 A

Fig. XXI, 9

presence or absence is to be determined (e.g., cadmium). Part of such a spectrum
is shown in Fig. XXI, 5. The graduated scale of wavelengths is also photographed
on the plate. Examination of the three spectra with a Rank Hilger projection
comparator will reveal at once the presence or absence of the specified element
(cadmium in the present example). This instrument, which is described in detail
above (Section XXI, 2), also enables comparisons to be made between images on
two different spectographic plates. Thus if a set of reference plates is available,
the need for taking comparison spectra of metals looked for against that of the
unknown is obviated and a great deal of tedious measurement is likewise
rendered unnecessary. Standard spectrographic samples are available inter alia
from the Bureau of Standards at Washington.

Mention may be made of the great advantage that the spectrograph offers in
the following operations :

(a) " ••

ib)

(c)

Rapid qualitative analysis of all the metallic constituents of a substance as
a basis for planning a chemical analysis.

Approximate analysis of minor components by sight (after some experience
has been obtained).

Examination of precipitates (after weighing) for freedom from constituents
(d) D separated.

election of traces of metallic impurities or constituents in inorganic
resi ues and powders ; in organic substances (foodstuffs, textiles, etc.) ; in
W T substances (glasses and slags) ; and in refractories and clays,

esting the purity of analytical reagents.

*'VithR.U.
and Cs.

powder two lines may be acceptable for some elements ; these include B, Cu, Au, P, Ag,

797

XXI, 4 QUANTITATIVE INORGANIC ANALYSIS

(/) Analysis of substances of which only small quantities are available.

(g) Detection of rare or trace metals in minerals.

XXI, 4. QUANTITATIVE SPECTRO GRAPHIC ANALYSIS. If the
excitation conditions are kept constant and the sample composition is varied
over a narrow range, the energy emitted for a given spectral line of an element is
proportional to the number of atoms that are excited, and thus to the
concentration of the element in the sample. The energy emitted (i.e., the intensity
of the light) is usually measured by the photographic method : the concentration
of the unknown is determined from the blackening of the plate for certain lines in
the spectrum. The quantitative determination of the blackening of the individual
lines is made with a microphotometer (Fig. XXI, 8). Measurements are made of
the light transmitted by the line in question (/) and the light transmitted by the
clear portion of the plate (/q), the density D (strictly the density of blackening,
also represented by B) may be defined by the expression D = logio(/o/0- It is
assumed that the galvanometer deflection obtained on the microphotometer, is
directly proportional to the light falling on the photocell.

The density of the image of the spectral line should ideally be proportional to
the concentration of the corresponding element in the sample if the exposure time
and conditions of excitation, etc., are held constant. This is often not strictly true ;
hence, wherever possible, it is desirable to photograph spectra of several samples
of varying known composition on the same plate with the unknown. The
unknown may then be evaluated by interpolation on a graph of density against
concentration. The fact that the unknown and all the standards are on the same
plate prevents errors due to differences in sensitivity between plates as well as
those due to differences in the time or temperature of photographic processing.

Since the intensity of the lines is ultimately registered on a photographic plate,
a brief discussion of the nature of the photographic process is desirable. If one
plots the density as a function of the logarithm of the exposure,* a curve such as is
shown in Fig. XXI, 10, results. A certain threshold exposure, denoted by A, is
necessary before an image is produced. It will be noted that there is a region BC
over which the density is proportional to the logarithm of the exposure ; this is the
useful range of the plate. The slope of this linear portion of the curve is known as
the gamma (or contrast) of the emulsion : y = tan 0. Emulsions with a high
gamma give images with strong contrast because a small difference in exposure
causes a large variation in density ; low values of gamma indicate low contrast.
The point D (the intercept of BC on the horizontal axis) measures the inertia of
the emulsion: its reciprocal is related to the ‘speed’ of the emulsion. For our
purpose the speed is an approximate measure of the minimum amount of light
required to produce a useful image. The slope of the characteristic curve of an
emulsion varies from emulsion to emulsion, with wavelength, and with the
conditions of excitation and of development. In order to determine the curve for
any given emulsion, the conditions of excitation (shape of the ends of the
electrodes, their distance apart, the electrical circuit, etc.) and also the conditions
of development (the type of developer, temperature, and time of development)
must be standardised. In selecting a plate on which to photograph the spectrum,

* The exposure of the photographic plate is defined as the product of the intensity of the light incident
on the undeveloped plate and the time for which it is acting.

798

EMISSION SPECTROGRAPHY XXI, 4

one decides first whether a fast emulsion is needed on the basis of the light
intensity available and the permissible time of exposure. If very faint spectrum
lines are to be detected, high sensitivity at low intensities is needed, which
suggests the use of a fast plate. To reproduce both weak and strong spectrum
lines on the same spectrogram with correct indication of their relative intensities,
medium contrast is needed. For sharp spectrum lines with a clear backgroimd, a
plate of high contrast is used. Slow contrast plates exhibit high resolving power :
such plates are of greatest value for use with spectrographs having low dispersion
with high resolving power as is often the case with certain prism instruments of
short focus. The most common type of plate used for emission spectrography has
a gamma of about 1.

We may now deal with some of the procedures employed in quantitative
spectrographic analysis. In the comparison sample method, the spectrum of an
unknown sample is compared with the spectra of a range of samples of known
composition (e.g., those supplied by the US Bureau of Standards) with respect to
a particular component or components. The spectra of the unknown and of the
various standards are photographed on the same plate under the same
conditions. The concentrations of the desired constituent can then be estimated
by coinparing the blackening of the lines of the particular constituent with the
same lines on the standards ; visual or photometric comparison of blackening
may be used.

In the internal standard method the intensity of the xmknown line is measured
relative to that of an internal standard line. The internal standard line may be a
weak line of the main constituent. Alternatively, it may be a strong line of an
element known not to be present in the sample and furnished by adding a fixed
small amount of a compound of the element in question to the sample. The ratios
of the intensities of these lines— the unknown line and the internal standard
line— will be unaffected by the exposure and development conditions. This
method will provide lines of suitable wavelength and intensity by variations of
t e added element and the amoimt added, due regard being paid to the relative
yo atihty of the selected internal standard element. It is important to use as
pairs only those lines of which the relative intensities are
variations in excitation conditions. The line selected as^ standard
I wavelength close to that of the unknown and should, if possible,
the same intensity.

insensitive to
should have i
have roughly

799

XXI, 5 QUANTITATIVE INORGANIC ANALYSIS

For initial experience in quantitative spectrographic analysis, procedures
involving solutions of materials are obviously attractive. The use of solutions has
the advantage that the constituents are uniformly distributed but the
disadvantage that it is not easy to ensure reproducible conditions for bringing the
solution into the light source (arc or spark). One procedure for obtaining spectra
from solutions is to add a small amount to spectroscopically pure carbon or
graphite electrodes and to arc or spark them after an initial drying period. The
solution technique allows a simple preparation of synthetic standards using
spectroscopically pure compounds, but it has certain defects ; thus elements, such
as silicon and tungsten, are difficult to keep in solution. The sensitivity of solution
methods is generally lower than that of other techniques.

XXI, 5. DIRECT READING INSTRUMENTS. A detailed discussion of
direct reading instruments is beyond the scope of this book as the apparatus is
expensive and only available in a few laboratories. However, a brief outline of the
principles of direct reading emission spectroscopy will be given.

A diagram of the light path in the Rank Hilger E950 Polyvac Direct Reading
Spectrometer is shown in Fig. XXI, 11. In this instrument the radiation is
dispersed by the holographic grating and the component wavelengths reach a
series of exit slits which isolate the selected emission lines for specific elements.
The light from each exit slit is directed to fall on the cathode of a photomultiplier
tube, one for each spectral line isolated. The light falling on the photomultiplier
gives an output which is integrated on a capacitor, thus the resulting voltage is a
function of the amount of element present in the sample. A calibration curve of
element concentration against capacitor voltage reading can be constructed.

Highly advanced direct reading instruments such as the Rank Hilger Polyvac

ElOOO offer a range of gratings, so
allowing a wide choice of spectral
wavelengths, from a 160nm to
864 nm (1600-8640 A). The com-
plete apparatus (with the excep-
tion of the excitation source) is
enclosed in a vacuum chamber in
order to permit adequate trans-
mission at wavelengths below
200 nm (2000 A) and to avoid
interference from the band due to
molecular oxygen at 180nm(1800
A). Thus elements such as carbon
and sulphur, which give rise to
emission lines in the vacuum
ultraviolet region, may be de-
termined.

The arc or spark excitation discharge is carried out in an argon atmosphere to
avoid spectral interference from the components of air. The main advantage of
direct reading instruments is that when used in conjunction with an on-line
computer they provide a method for the rapid simultaneous analysis of elements
with a better precision and accuracy than can be normally obtained from the
spectrograph using a photographic plate. Thus results for 25 elements may be
obtained within a short time of about 1-2 minutes.

800

EMISSION SPECTROGRAPHY XXI, 6

Experimental

XXI 6. QUALITATIVE SPECTROGRAPHIC ANALYSIS* OF (A) A

non-ferrous alloy and (B) a complex inorganic

MIXTURE. General discussion. One procedure for the identification of an
alloy is to measure the wavelength values of the observed lines and compare these
with the recorded data on known elements (see table of persistent lines in
Appendix XIV). A wavelength scale, which has been calibrated with the
spectrograph, is photographically reproduced on the plate: this is, of course,
only useful as a guide, since the wavelengths cannot be read with sufficient
accuracy; A simple method for use with brass, or most other simple non-ferrous
alloys, is to determine the spectra with pure samples of the component metals and
to compare the spectra by a projection method (see Fig. XXI, 6). At least three
persistent lines must be present for positive identification. The spark technique
may be used for metals and alloys, but is not altogether satisfactory for powders,
including R.U. powder. The d.c. carbon arc is preferred for the qualitative
analysis of powders, and it also gives good results for most alloys ; low-melting
alloys (e.g.. Wood’s metal) may be dissolved in nitric acid, evaporated to dryness,
then evaporated with concentrated sulphuric acid, and the dry sulphated residue
employed in the carbon arc.

To gain experience in qualitative analysis, full details will be given for the
analysis of brass and an artificial seven radical inorganic mixture.

Adjustment of the optical system. The condensing lens is set between the
light source and the slit of the spectrograph so that the beam of light from the d.c.
arc source passing through the slit forms a real image at the collimating lens
(compare Fig. XXI, 1) of the spectrograph. The adjustments for a Hilger and
Watts Medium Spectrograph will be evident from Fig. XXI, 12.

1. Place the Gramont stand so that the electrodes are 38 cm from the jaws of
the slit and align them for height ; the gap between the electrodes to be 4 mm.

Gauge

plate

Light

source
Lens I

Lens Slit

I 1

Collimator

lens

Spectrograph-I

Fig. XXI, 12

2. Set the condensing lens in position at 2 cm from the slit jaws and at the
correct height.

3. Set the gauge plate on its stand at the end of the bar at about 23 cm from the

qualitative and quantitative analysis described utilise a Hilger and Watts
instn ..^F^ctrograph. They can easily be adapted to other similar spectrographs with the aid of the
c ion manuals supplied by the manufacturers of the instruments.

801

XXI, 6 QUANTITATIVE INORGANIC ANALYSIS

lens on the Gramont stand. An image of the light source will now be seen on the
gauge plate, and the height should be set so that the image agrees with the two
outer horizontal lines on the gauge. If a reading lamp is so arranged that it can be
brought into position near the electrodes on the slit side of the stand it will be
found that an image of the electrodes can be seen on the screen of the gauge plate.

Thus for any subsequent work the electrodes can be set in
optical alignment with the spectrograph without actually
passing any current between them. The two outer
horizontal lines of the gauge are equivalent to an
elelctrode gap of 4 mm and the inner lines to a 2-mm gap.

d.c. arc. A 230-volt arc at 4 amperes is suitable for
qualitative analysis (see Fig. XXI, 3). The arc gap may be
2 mm and the slit width 0.02 mm.

Electrodes for d.c. arc. The two electrodes are
shown in Fig. XXI, 13. They are conveniently shaped on a
lathe from graphite electrodes (Johnson, Matthey ; 30 cm
tong; JM 3B, 10 mm diameter; JM 4B, 6.5 mm diameter).
The maximum depression on the lower electrode is 3 mm :
the small projection in the centre helps to ensure that the
arc passes between it and the upper electrode and does not ‘wander’ appreciably
to the edges of the electrode. A small quantity (about 20 mg) of.the alloy or
powder is placed on the lower electrode.

Photographic details. Dark room. A fully equipped dark room is
desirable. Ideally the spectrograph should be set up in the dark room. The
student should become familiar with its facilities— stainless-steel trays
(26 cm X 21 cm), running water, safelights, etc.

Plates. Ilford R.40 or Kodak V-F. Charge holder in dark room.

Photographic developer. Ilford ID-2* or equivalent. Dilute 1 volume with
2 volumes of water.

Photographic fixer. Kodak tropical acid hardening-fixing bath F5t or
equivalent.

Procedure, (a) Load the plate holder with plate, making certain that the
sensitised side is placed face down in the open holder. Return the holder to the
spectrograph.

Withdraw the safety slide. Expose for several seconds to obtain the wavelength
scale in the upper part of the plate, replace slide.

(b) Charge the lower electrode (anode) with about 20 mg of brass or the
inorganic mixture.

(c) Withdraw the safety slide which covers the plate. Strike arc between the
electrodes. Make an exposure of 6 seconds by opening the shutter of the
spectrograph with the Hartmann diaphragm in position. (The best exposure time
is evaluated by making a number of consecutive exposures, the plate being
lowered to the next position after each exposure.)

6-5 mm

Depth of
depression 3mm

• 10 mm

Fig. XXI, 13

• This has the following composition: melol, 2g; sodium sulphite (anhydrous) 75 g; hydroquinone,
8 g; sodium carbonate (anhydrous), 37.5 g; potassium bromide, 2 g; water to make I dm^. Dissolve
chemicals in order given. {Ilford Technical Information Book, Sheet D.20.1, Vbl. Ill, 1971 issue.)
t This has the following composition: sodium thiosulphate, 240.0g; sodium sulphite (crystals),
30.0 g; acetic acid (glacial), 17.0cm^: boric acid 7.5 g; potash alum 15.0g; water to make 1 dm^.
Dissolve chemicals in order given. {Kodak Data Book on Photography. Data Sheet W. 1 8.)

802

XXI, 7 QUANTITATIVE INORGANIC ANALYSIS

alloys provided a suitable internal standard can be found, e.g., lead, cadmium,
and copper in zinc using selected zinc lines as internal standards.

Preparation of solutions. Use spectroscopically standardised substances
(Johnson, Matthey) throughout, and also analytical reagent acids.

Copper solution. Dissolve about 8.81 g, accurately weighed, of copper* sheet in
i:i-nitric acid (about 60 cm^) and dilute to lOOcm^ with distilled water in a
graduated flask.

Magnesium solution. Dissolve about 0.88 g, accurately weighed, of magnesium
in 10 per cent nitric acid (1:9 v/v), and dilute to 500 cm^ in a graduated flask with
10 per cent nitric acid.

Lead solution. Dissolve about 1.00 g, accurately weighed, of lead nitrate in the
minimum volume of water and dilute to 250 cm^ in a graduated flask with 1 0 per
cent nitric acid.

Standard solutions. Prepare the standard solutions by mixing the following
volumes of each of the above solutions and diluting each to 25 cm^ with distilled
water in a graduated flask.

standard Volumes of solutions, cm^ Pb

solution concentration,*

Cu Mg Pb mg per cm-’

10.00

5.00

1.50

0.1506

10.00

5.00

2.00

0.2008

10.00

5.00

3.00

10.00

5,00

4.00

0.4016

10.00

5,00

5.00

♦ These figures apply to 8.8150g Cu, 0.8861 g Mg, and 1.0088 g PblNOjlj.

Brass solution. Weigh out accurately about 5.0 g of a standard brass containing
about 2 per cent Pb, treat with 1 : 1 -nitric acid (about 40 cm^) until it is completely
attacked, and warm on a water bath for 30 minutes. Dilute with a little water,
filter (Whatman paper No. 541), and wash the residue with three 10 cm^ portions
of hot distilled water. Transfer the combined filtrate and washings to a 100 cm^
graduated flask and dilute to the mark with 10 per cent nitric acid.

In order to test the method over a fairly wide concentration range, the brass
solution (prepared as above) may be diluted with the standard stock solutions as
follows. The final volume is made up with water.

Brass solution.

Cu solution.

Mg solution,

Total final

cm^

volume, cm^

Solution 6

5.00

8.00

5.00

25.0

Solution 7

5.00

6.00

4.00

20.0

Solution 8

5.00

4.00

3.00

15.0

These diluted brass solutions contain 0,2-0. 4 mg of Pb cm the copper
concentration of each solution is about the same (35.3 mg cm “ ^).

‘The concentration of the ‘main element in the standard and unknown solutions should be
approximately equal. The concentration of the internal standard should be constant in all the
solutions.

804

EMISSION SPECTROGRAPHY XXI, 7

Apparatus. Using the Hilger medium spectrograph, set up the optical
system as in Fig. XXI, 14. Place the short focus quartz lens B so that an image of

Collimator

1 ®

^ El ^ T

' Lens

) f* 0

sill 1— Spectrograph

reducing
■ wedge

Light Aperture Lens

source screen

Fig. XXI, 14

the electrode A appears on the aperture screen C; the last-named is a vertical
metal screen provided with a horizontal slit about 4 mm wide. Arrange the arc
gap (cn. 5 mm) so that the image of the tips of the electrodes lies just outside the
gap in the screen ; if desired, marks may be made on the screen so as to ensure a
constant arc gap. By this means radiation from the incandescent electrodes can
be excluded from the spectrograph and the entire positioning of the arc image
viewed and controlled during the exposure; the ‘background’ on the spectrum
caused by incandescence at the tips of the electrodes is largely eliminated. Place
the longer focus lens D close to the slit F : an image of the aperture is focused by-
the collimator lens on the prism.

The slit-reducing wedge and Hartmann diaphragm E is shown in Fig. XXI, 15.
This consists of a steel slide which fits into grooves in the face of the slit.
Movement of the wedge controls the length of the slit. By means of the Hartmann
diaphragm, generally used for qualitative analysis (see Section XXI, 6), several

0-5mm

-i

|-^ 0-5 mm
6mm

3 mm
6-5 mm

Fig. XXI, 16

spectrograms can be recorded on the same negative without moving the plate
holder. When the slide is moved horizontally any of several apertures allows the
light from the source to strike the desired portion of the slit; by appropriate
positioning of this device, comparison spectra above and below the unknown can
be photographed.

Shape the graphite electrodes (Johnson, Matthey, 4B) as shown in Fig.

XXl, 16, with the aid of a lathe. Produce the d.c. arc as described in Section
XXI, o.

805

XXI, 7 QUANTITATIVE INORGANIC ANALYSIS

Use Ilford N30, Kodak III-O or Agfa 34.B.50 plates. The other photographic
details are as in the previous section.

Procedure. By means of a micro-pipette, transfer 0.05 cm^ of each solution
into the cavity of an electrode; when this volume has been absorbed introduce a
further 0.05 cm®. Use a separate electrode for each solution. When absorption in
the walls of the crater is complete, dry the electrodes in an oven at 1 10° for 30-60
minutes.

Mount a pair of electrodes in the arc stand ; the standard solution is absorbed
in the lower (positive) electrode. Strike the arc, and photograph the spectrum for
an appropriate time immediately the arc is struck. Allow the electrodes to cool,
change both the upper and lower electrodes (the latter charged with another
standard solution), and repeat the exposure using another portion of the same
plate. Repeat with the remaining standard solutions and the various ‘unknown’
brass solutions.

Develop the plate using ID-2 developer for 3.5 minutes at 16°C, fix using
Kodak tropical acid hardening-fixing bath F5 (or equivalent) for 10-15 minutes,
wash for 10-15 minutes, rinse with distilled water, and allow to dry.

Employ a Hilger non-recording microphotometer (Fig. XXI, 8) to determine
the blackening of the appropriate lines on the spectrogram with due regard to the
following points :

(fit) The emulsion side of the plate must face the light source.

(h) The slit width should be just less than ten times the slit width of the
spectrograph (a slit width of 0. 1 2 mm was found to be satisfactory).

(fiO The reference lines should lie close together, e.g.. Mg 2776.7 or 2783.0 A and
Pb 2873.3 A.

(e) The clear portion of the plate to be illuminated should be as near as possible
to the line to be measured.

Tabulate the results in the form shown below for some typical values obtained
for this determination. Plot log (‘Density’ Pb/‘Density’ Mg) against log
(Concentration of Pbx 10). From the resulting straight-line plot, evaluate the
lead concentrations for solutions 6, 7, and 8 (these results are marked with an
asterisk in the table). The weights of brass per cm® in solutions 6, 7, and 8 are
known, and hence the percentage of Pb in the samples of brass can be calculated.

Typical results for the spectrographic analysis for lead (internal standard: magnesium)

Solution

Densitometer reading

. / ‘Density’ Pb \

1 ‘Density’ Mg |

Concn. of

Log (concn.

number

Pb, mgcm'^

ofPbxlO)

7.3

17.5

0.380

0.178

8.7

18.5

0.328

7.3

11.0

0.178

0.3012

0.479

10.3

12.0

0.066

0.4016

0.604

11.6

10.4

-0.047

8.8

15.5

0.246

0.251*

9.2

12.5

0.133

0.335*

10.7

12.2

0.057

0.612*

The results were; for solution 6, 2.51 per cent; for solution 7, 2.67 per cent; for
solution 8, 2.45 per cent, thus giving a mean Pb content of 2.54 per cent with a

806

EMISSION SPECTROGRAPHY XXI, 8

standard deviation ±0.12 per cent. The true value for thesample of brass used
was 2.52 per cent. ■ ' ' , ' . :

vvT « nFTFRMINATION OF COPPER AND LEAD IN WHITE
METAL. This experiment has been designed for students with the
nrimary objects :.(i) to illustrate the use of internal standards m the quanUtat
soectroCTaphic analysis of an alloy for two elements, and (u) to P^ov
eLerience^in the use of a non-recording microphotometer; The al oy selected

(wLe metal) is readily available, and the two elements (copper ^

elsily determined by purely chemical methods. It is appreciated that, the
percentages™ the two dements (about 4 per cent each) present m the alloy are
^eTmuch higher than would normally be determined by spe-tro^ic
methods (which are generally confined to percentages less
nevertheless the experiment will indicate the upper limits possible and the
accuracy attainable; the latter may be regarded secondary objects of the
exercise. With the experience so gamed, the student should be able to adapt the
procedure to the determination of two or more elements present m proportions
( < 1 per cent) for which spectrographic techniques are eminently suitable. 1 he
approximate proportions of the elements to be determined must be known m

order to prepare the standard solutions. , l ..

Preparation of solutions. Use spectroscopically standardised substances

(Johnson, Matthey) throughout; also analytical reagent acids. ' ■

Copper solution. Dissolve about 1.62 g, accurately weighed, of copper^ in a
mixture of about 50 cm^ concentrated hydrochloric acid and 5-1 0 cm
concentrated nitric acid, boil the solution for a few minutes, cool, and dilute to
500 cm^ with hydrochloric acid (3.5 :i,v/v). , r •

MagHesiwH solution. Dissolve about 1.05 g, accurately weighed, of magnesium
in 50 cm^ concentrated hydrochloric acid and dilute to 500 cm^ with hydro-
chloric acid as above. • n 3

Tin solution. Dissolve about 10.5 g, accurately weighed, of tin m 60 crn
concentrated hydrochloric acid, and dilute to 100 cm^ with hydrochloric acid
(i:i,v/v).

Lead solution. Dissolve about 2.24 g, accurately weighed, of lead nitrate in
10 cm^ water and add about 100 cm^ concentrated hydrochloric acid: boil until
the lead chloride dissolves (5-10 minutes), cool, aiid dilute to 500 cm^ with
concentrated hydrochloric acid.

Standard solutions. Prepare the standard solutions by mixing the following
volumes of each of the above solutions and diluting each tp .25cm^ with
concentrated hydrochloric acid in a graduated flask.

Standard

Volumes of solutions, cm

solution

concentration.

mg cm~^*

10.00

5.00

1.00

0.13

0.11

10.00

5.00

2.00

0.26

0.22 ,

10.00

5.00

3.00

0.39

0.34 '

10.00

5.00

4.00

0.53

0.45

10.00

5.00

0.66

0.56

* These are approximate values; the exact concentrations will depend upon the weights used in the
preparation of the various solutions. , .

807

XXI, 9 QUANTITATIVE INORGANIC ANALYSIS

White metal solution. Weigh out accurately about 5.18 g of the white metal
alloy, and add about 50 cm^ concentrated hydroehloric acid (vigorous reaction).
Treat the suspension drop wise with concentrated nitric acid until a clear green
solution results, boil gently to remove nitrous fumes, cool, and make up to
100 cm^ with concentrated hydrochloric acid in a graduated flask. Dilute the
alloy solution with known amounts of tin solution so that the resulting solutions
contain 0.30-0.60 mg copper cm”^ and 0.25-0.55 mg lead cm"^: magnesium
solution must be added in amounts to ensure that its concentration is almost
identical with that in the standard solutions. The final volumes are made up with
concentrated hydrochloric acid added from a burette.

White metal
solution, cm ^

Sn solution,
cm^

Mg solution,
cm^

Total final
volume, cm^

Solution 6

5.00

6.00

4.00

20.0

Solution 7

5.00

8.00

5.00

25.0

Solution 8

5.00

10.00

6.00

30.0

Procedure. Follow the method given in Section XXI, 7. Introduce three
separate portions of 0.05 cm^ into the electrodes.

Use the following reference lines: Mg, 2776.7 A; Cu, 2824.4 A; and Pb,
2873.3 A. Measure the densities of the lines with the Hilger non-recording
raicrophotometer.

Plot log (‘Density’ Pb/‘Density’ Mg) against log (Concentration of Pb x 10),
and log (‘Density’ Cu/‘Density’ Mg) against log (Concentration of CuxlO).
Evaluate the Pb and Cu concentrations for Solutions 6, 7, and 8, and thence the
corresponding percentages of Pb and Cu in the sample of white metal. Some
typical results for Pb were 3.74, 3.60, and 3.84 per cent.

XXI, 9. Selected bibliography

1. W. R. Erode (1943). Chemical Spectroscopy. 2nd edn. New York; John Wiley.

2. S. Judd Lewis (1946). Spectroscopy in Science and Industry. 2nd edn. London;
Blackie.

3. L. N. Ahrens and S. R. Taylor (1960). Spectrochemical Analysis. 2nd edn. Reading,
Mass. ; Addison- Wesley.

4. J. Sherman (1960). Emission Specirography, in W. G. Berl, Physical Methods of
Chemical Analysis. Vol. 1. 2nd edn. New York; Academic Press.

5. F. Twyman (1951). Metal Spectroscopy. London; Charles Griffin.

6. C. R. N. Strouts, H. N. Wilson and T. R. Parry-Jones (1962). Chemical analysis.
The Working Tools. London; Oxford University Press.

7. G. W. Ewing (1975). Instrumental Methods of Chemical Analysis. 4th edn. New York;
McGraw-Hill Book Co.

8. H. A. Strobel (1973). Chemical Instrumentation. A Systematic Approach to
Instrumental Analysis. 2nd edn. Reading, Mass. ; Addison- Wesley Publishing Co.

9. A. N. Zaidel et al. (1970). Tables of Spectral Lines. 3rd edn. New York; Plenum
Publishing Co.

10. M. Slavin (1971). Emission Spectrochemical Analysis (Chemical Analysis Series). New
York; Wiley-Interscience.

808

EMISSION SPECTROGRAPHY XXI, 9

11. H. H. Willard, L. L. Merritt and J. A. Dean (1974). Instrumental Methods of Analysis.
5th edn. New York; Van Nostrand-Reinhold.

12. I. M. KolthoffandP. J. Elving(eds.)(1965). Treatise on Analytical Chemistry, Part 1,
Vol. 6, Ch. 64. B. F. Scribner and M. Margoshes, Emission Spectroscopy. New York;
Interscience.

CHAPTER XXII FLAME SPECTROMETRY

xxn, 1. GENERAL DISCUSSION. If a solution containing a metallic salt
(or some other metallic compound) is aspirated into a flame (e.g., of acetylene
burning in air), a vapour which contains atoms of the metal may be formed.
Some of these gaseous metal atoms may be raised to an energy level which is
sufficiently high to permit tfiTemission of radiation characteristic of that metal;
e.g., the characteristic yellow colour imparted to flames by compounds of
sodium. This is the basis of flame emission spectroscopy (FES) which was
formerly referred to as flame photometry. However, a much larger number of the
gaseous metal atoms will normally remain in an unexcited state or, in other
words, in the ground state. These ground state atoms are capable of absorbing
radiant energy of their own specific resonance wavelength, which in geileral is the
wavelength of the radiation that the atoms would emit if excited from the ground
state. Hence if light of the resonance wavelength is passed through a flame
containing the atoms in question, then part of the light will be absorbed, and the
extent of absorption will be proportional to the number of ground state atoms
present in the flame. This is the underlying principle of atomic absorption
spectroscopy (AAS). Atomic fluorescence spectroscopy (AFS) is based on the re-
emission of absorbed energy by free atoms.

The procedure by which gaseous metal atoms are produced in the flame may be
summarised as follows. When a solution containing a suitable compound of the

M*X~ ^

Solution

M’gas)

Thermal

Excitation

Flame | hv
emission \

Fig. xxn, 1

Mist

Htgas)"*" X(gas)

Evaporation

Solid

Dissociation

Vapourisation

Gas

Absorption of .Radiant Energy

'i'

M-gos)

Re-emission (Huorescence)

hv or hv'

810

FLAME SPECTROMETRY XXH, 2

metal to be investigated is aspirated into a flame, the following events occur in
rapid succession : ■

1 . evaporation ofsolvent leaving a solid residue; ^

2. vaporisation of the solid with dissociation into its constituent atoms, which
initially, will be in the ground state;

3. some atoms may be excited by the thermal ener^ of the flame to higher energy

levels, and attain a condition in which they radiate energy.

The resulting emission spectrum thus consists of lines originating from excited
atoms or ions. These processes are conveniently represented diagrammatically as
in Fig. XXII, 1 .

xxn, 2 . ELEMENTARY THEORY. Consider the simplified energy level
diagram shown in Fig. XXII, 2 , where Eq represents the ground state in which the
electrons of a given atom are at their lowest energy level and E^, E2, E^, etc.,
represent higher or excited energy levels.

Transitions between two quantised energy levels, say from £"0

correspond to the absorption of radiant energy, and
the amount of energy absorbed (AE) is determined
by Bohr’s equation

AE = £1 — £(, = /2V = hc/X

where c is the velocity of light, h is Planck’s
constant, and v is the frequency and L the
wavelength of the radiation absorbed. Clearly, the
transition from £4 to Eq corresponds to the emission
of radiation of frequency v.

Since an atom of a given element gives rise to a
definite, characteristic line spectrum, it follows that
there are different excitation states associated with
different elements. The consequent emission spectra involve not only transitions
from excited states to the ground state, e.g., £3 to Eq, £2 to Eq (indicated by the
bold lines in Fig. XXII, 2 ), but also transitions such as £3 to £3, £3 to £j, etc.
(indicated by the dotted lines). Thus it follows that the emission spectrum of a
given element may be quite complex. In theory it is also possible for absorption of
radiafion by already excited states to occur, e.g., Ey to £3, £3 to £3, etc., but in
practice the ratio of excited to ground state atoms is extremely small, and thus the
absoiption spectrixm of a given element is usually only associated with
transitions from the ground state to higher, energy states and is consequently
much simpler in character than the emission spectrum. ,

The relationship between the ground state and excited state populations is
given by the Boltzmann equation

N,/No = (gj/go)e-“?'=^

where = number of atoms in the excited state,

No = number of ground state atoms,
gi/go = ratio of statistical weights for ground and excited states
AE = energy of excitation = hv, . ’

k = the Boltzmann constant,
r= the temperature in Kelvin.

811

XXn, 2 QUANTITATIVE INORGANIC ANALYSIS

It can be seen from this equation that the ratio Nj/Nq is dependent upon both the
excitation energy and the temperature T. An increase in temperature and a
decrease in AE (i.e., when dealing with transitions which occur at longer
wavelengths) will both result in a higher value for the ratio Nj/Ng.

Calculation shows that only a small fraction of the atoms are excited, even
under the most favourable conditions, i.e., when the temperature is high and the
excitation energy low. This is illustrated by the data in Table XXII, I for some
typical resonance lines.

Table XXII, 1 Variation of atomic excitation with wavelength and with temperature

Element

Wavelength (nm)

N,/N„

2000 K

4000 K

589.0

9.86x10-“

4.44x10-“

422.7

1.21 X 10-'’

6.03 X 10-*

213.9

7.31 X 10-'“

1.48 xlO-'’

Since as already explained the absorption spectra of most elements are simple
in character as compared with the emission spectra, it follows that atomic
absorption spectroscopy is less prone to inter-element interferences than is flame
emission spectroscopy. Further, in view of the high proportion of ground state to
excited atoms^it would appear that atomic absorption spectroscopy should also
be more sensitive that flame emission spectroscopy. However, in this respect the
wavelength of the resonance line is a critical factor, and elements whose
resonance lines are associated with relatively low energy values are more sensitive
as far as flame emission spectroscopy is concerned than those whose resonance
lines are associated with higher energy values. Thus sodium with an emission line
of wavelength 589.0 nm shows great sensitivity in flame emission spectroscopy,
whereas zinc (emission line wavelength 213.9 nm) is relatively insensitive.

The integrated absorption is given by the expression

Xdv = fN^ine^jmc)

where

K is the absorption coefficient at frequency v,
e is the electronic charge,
m the mass of an electron,
c the velocity of light,

/ the oscillator strength of the absorbing line (this is inversely
proportional to the lifetime of the excited state),

Nq is the number of metal atoms per cm^ capable of absorbing the
radiation.

In this expression the only variable is and it is this which governs the extent of
absorption. Thus it follows that the integrated absorption coefficient is directly
proportional to the concentration of the absorbing species.

It would appear that measurement of the integrated absorption coefficient
should furnish an ideal method of quantitative analysis. In practice, however, the
absolute measurement of the absorption coefficients of atomic spectral lines is
extremely difficult. The natural line width of an atomic spectral line is about
10 ^ nm, but owing to the influence of Doppler and pressure effects, the line is

812

FLAME SPECTROMETRY XXn,2

broadened to about 0.002 nm at flame temperatures of 2000-3000 K. To measure
the absorption coefficient of a line thus broadened would require a spectrometer
with a resolving power of 500000. This difficulty was overcome by Walsh (Ref.
1), who used a source of sharp emission lines with a much smaller half width than
the absorption line, and the radiation frequency of which is centred on the
absorption frequency. In this way, the absorption coefficient at the centre of the
line, may be measured. If the profile of the absorption line is assumed to be

due only to Doppler broadening, then there is a relationship between and
No- Thus the only requirement of the spectrometer is that it shall be capable of
isolating the required resonance line from all other lines emitted by the source.

It should be noted that in atomic absorption spectroscopy, as with molecular
absorption, the absorbance A is given by the logarithmic ratio of the intensity of
the incident light signal /q to that of the transmitted light /(, i.e.,

A = log IqI It = KLNq

where Nq is the concentration of atoms in the flame (number of atoms per
cm^);

L is the path length through the flame (cm),

Kisa constant related to the absorption coefficient.

For small values of the absorbance, this is a linear function.

With flame emission spectroscopy, the detector response E is given by the
expression

E = kac,

where ^ris related to a variety of factors including the efficiency of
atomisation and of self absorption,

a is the efficiency of atomic excitation,
c is the concentration of the test solution.

It follows that any electrical method of increasing E, as for example, improved
amplification, will make the technique more sensitive.

The basic equation for atomic fluorescence is given by

F= Qlokc

where Q is the quantum efficiency of the atomic fluorescence process,

/q is the intensity of the incident radiation,
kis a constant which is governed by the efficiency of the
atomisation process,

c is the concentration of the element concerned in the test solution.

It follows that the more powerful the radiation source, the greater will be the
sensitivity of the technique.

To summarise, in both atomic absorption spectroscopy and in atomic
fluorescence spectroscopy, the factors which favour the production of gaseous
atoms in the ground state determine the success of the techniques. In flame
enussion spectroscopy, there is an additional requirement, namely, the
production of excited atoms in the vapour state. It should be noted that the
conversion of the original solid MX into gaseous metal atoms (M ) will be
governed by a variety of factors including the rate of vapourisation flame
composition and flame temperature, and further, if MX is replaced by a new

813

XXII, 3/4 QUANTITATIVE INORGANIC ANALYSIS

solid, MY, then the formation of A/ga^ may proceed in a different manner,- and
with a different efficiency from that observed with MX.

XXn, 3, INSTRUMENTATION. The three flame spectrophotometric
procedures require the following essential apparatus.

(a) For flame emission spectroscopy a nebuliser-burner system which produces
gaseous metal atoms by using a suitable combustion flame involving a fuel
gas-oxidant gas mixture is needed. Note however that with so-called non-flame
cells, the burner is not required.

(b) A spectrophotometer system which includes a suitable optical train, a
photosensitive detector and appropriate display device for the output from the
detector.

(c) For both atomic absorption spectroscopy and atomic fluorescence
spectroscopy, a resonance line source is required for each element to be
determined ; these line sources are usually modulated (see Section XXII, 9).

A schematic diagram showing the disposition of these essential components
for the different teehniques is given in Fig. XXII, 3. The components included

Fig.xxn,3

within the frame drawn in dashed lines represent the apparatus required for flame
emission spectroscopy. For atomic absorption spectroscopy and for atomic
fluorescence spectroscopy there is the additional requirement of a resonance line
source. In atomic absorption spectroscopy this source is placed in line with the
detector, but in atomic fluorescence spectroscopy it is placed in a position at right
angles to the detector as shown in the diagram. The essential eomponents of the
apparatus required for flame spectrophotometric techniques will be considered
in detail in the following sections.

XXII, 4. COMBUSTION FLAMES. For flame spectroscopy an essential
requirement is that the flame used shall produce temperatures in excess of
2000 K. In most cases this requirement can only be met by burning the fuel gas in
an oxidant gas which is usually air, nitrous oxide, or oxygen diluted with either
nitrogen or argon. The flame temperatures attained by the common fuel gases
burning in (i) air and (ii) nitrous oxide are given in Table XXII, 2 ; the value given
for town gas/air can only be regarded as approximate since it will depend upon

814

FLAME SPECTROMETRY XXII, 5

the exact composition of the ‘town gas’. The flow rates of both the fuel gas and
the oxidant gas should be measured, for some flames are required to be rich m the
fyel gas, whilst other flames should be lean in fuel gas ; these requirements are
discussed in Section XXH, 20. The concentration of gaseous atoms within the
flame, both in the ground and in the excited states; may be influenced by (a) the
flame composition, and by (6) the:position considered within the flame.

Table XXII, 2 Flame temperatures Mth various fuels

Fuel gas

Temperature (T/K)

Air’

Nitrous oxide

Acetylene

2400

3200

Hydrogen

2300

2900

Propane

2200

3000 .

Town gas

2100

—

As far as flame composition is concerned, it may be noted that an acetylene-air
mixture is suitable for the determination of some thirty metals, but a propane-air
flame is to be preferred for metals which are easily converted into an atomic
vapour state. For metals such as aluminium and titanium which form refractory
oxides, the higher temperature of the acetylene-nitrous oxide flame is essential,
and the sensitivity is found to be enhanced if the flame is fuel rich.

With regard to position within the flame, it can be shown that in certain cases
the concentration of atoms may vary widely if the flame is moved either vertically
or laterally relative to the light path from the resonance line source. Rann and
Hambly (Ref. 2) have shown that with certain metals (e.g., calcium and
molybdenum), the region of maximum absorption is restricted to specific areas of
the flame, whereas the absorption of silver atoms does not alter appreciably
within the flame, and is unaffected by the fuel gas/oxidant gas ratio.

For the sake of brevity, the so-called ‘cool flame’ techniques based upon the
use of an oxidant-lean flame such as hydrogen/nitrogen-air, have not been
included. Details can however be found in Ref. 13, and it should also be noted
that the experiment described in Section XXII, 26 utilises a ‘cool flame’.

XXn, 5. THE NEBULISER-BURNER SYSTEM. The purpose of the
nebuliser-bumer system is to convert the test solution to gaseous atoms as
indicated in Fig. XXII, i, and the success of flame photometric methods is
dependent upon the correct functioning of the nebuliser-bumer system. It should
however, be noted that some flame photometers have a very simple burner
system (see Section XXII, 12).

The furiction of the nebuliser is to produce a mist or aerosol of the test solution.
The solution to be nebulised is drawn up a capillary tube by the Venturi action of
a jet of air blowing across the top of the capillary; a gas flow at high pressure is
necessary in order to produce a fine aerosol.

There are two main types of burner system : (a) the Pre-mix or Laminar-flow
burner, and (b) the Total Consumption or Turbulent-flow burner. In the Pre-mix
of burner, the aerosol is produced in a vapourising chamber where the larger
droplets of liquid fall out from the gas stream and are discharged to waste. The
resulting fine mist is mixed with the fuel gas and the carrier (oxidant) gas, and the

815

XXn, 5 QUANTITATIVE INORGANIC ANALYSIS

mixed gases then flow to the burner head. In atomic absorption spectroscopy the
burner is a long horizontal tube with a narrow slit along its length. This produces
a thin flame of long path length which can be turned into or away from the beam
of radiant energy. The flame path of a burner using air-acetylene, air-propane or
air-hydrogen mixtures is about 1 0-1 2 cm in length, but with a nitrous
oxide-acetylene burner it is usually reduced to about 5 cm because of the higher
burning velocity of this gas mixture. In addition to a long light path, this type of
burner has the advantages of being quiet in action and with littl e danger of
incrustatio n around the burner head smcrlaf aeTlropfets of solut loiLhave been
elimrnatea~from th^treanrof'gasTeachi i^lhe burner. Its disadvantages are (i)
that' with-solutionsTnade'Up"ifnmxeaioIvents, the more volatile solvents are
evaporated preferentially ,- (ii) a potential explosion hazard exists since the burner
uses relatively large volumes of gas, but in modern versions of this type of
burner this hazard is minimised.

A typical burner of this type is shown in Fig. XXII, 4. In this particular burner
(Perkin-Elmer Corporation), the mixing chamber is a steel casting lined with a

plastic (‘Penton’) which is extremely resistant to corrosion. The burner head is
manufactured from titanium, thus avoiding the occasional high readings which
are encountered when solutions containing iron and copper in presence of acid
are examined with burners having a stainless steel head. The nebuliser is capable
of adjustment so that it can handle sample up-take rates of from 1-5 cm^ per
minute. The burner can be adjusted in three directions, and horizontal and
vertical scales are provided so that its position can be recorded. The head may be
turned through an angle of 90° with respect to the light beam, and so the path
length of the flame traversed by the resonance line radiation may be varied
considerably: by choosing a small path length it becomes possible to analyse
solutions of relatively high concentration without the need for prior dilution.

The Total Consumption type of burner consists of three concentric tubes

816

FLAME SPECTROMETRY XXH, 6

shown in Fig. XXII, 5. The sample solution is oypdeclbiLaJne^piUari tube A
directly into the flame. The fuel gas-anarE ^fdant gas a recarnedaloM|gparate
that they only mi}rat'tlie Tit£g Jfie'burner . Since alTtp liquid samj)le
■ k aspiratpdii p j ^th^apiflam ube^^ fl ame, it would appear that

this type of burne? should be more efficient thanlEe pre-mix type of bumen
However, the total consumption burner gives a flame of relatively s^rt_p^h
length, and hence such burners are predominantly used for flame enussion
studied This type of burner has the advantages that (i) it is simple to
manufacture, (ii) it allows a totally representative sample to reach the flame, and
(iii) it is free from explosion hazards arising from unburnt gas mixtures. Its

Fuel gas inlet
Carrier gas

1 1

k Sample
' inlet

Fig.XXn,5

disadvantages are that (i) the aspiration rate varies with different solvents, and
(ii) there is a tendency for incrustations to form at the tip of the burner which can
lead to variations in the signal recorded.

In general terms, Thomerson and Thompson (Ref. 3) have cited the following
disadvantages of flame atomisation procedures :

1 . Only 5-1 5 per cent of the nebulised sample reaches the flame (in the case of the
pre-mix type of burner) and it is then further diluted by the fuel and oxidant
gases so that the concentration of the test material in the flame may be
extremely minute.

2. A minimum sample volume of between 0.5 and 1.0 cm^ is needed to give a
reliable reading by aspiration into a flame system.

3. Samples which are viscous (e.g., oils, blood, blood serum) require dilution
with a solvent, or alternatively must be ‘wet ashed’ before the sample can be
nebulised.

XXn, 6. NON-FLAME TECHNIQUES. Instead of employing the high
temperature of a flame to bring about the production of atoms from the sample,
it is possible in some cases to make use of either (a) non-flame methods involving
the use of electrically heated graphite tubes or rods, or (b) vapour techniques.
Procedures (a) and (b) both find applications in atomic absorption spectroscopy
and in atomic fluorescence spectroscopy.

(a) Electrothermal atomisers, (i) The graphite tube furnace. A diagram
of a graphite tube furnace is shown in Fig. XXII, 6. It consists of a hollow
graphite cylinder about 50 mm in length and about 9 mm internal diameter and
so situated that the radiation beam passes along the axis of the tube. The graphite

817

XXII, 6 QUANTITATIVE INORGANIC ANALYSIS

Removable Water

window in

Insulator

Electrical

connector

Water out

Fig. XXII, 6

tube is surrounded by a metal jacket through which water is circulated and which
is separated from the graphite tube by a gas space. An inert gas, usually argon, is
circulated in the gas space, and enters the graphite tube through openings in the
cylinder wall.

The solution of the sample to be analysed (1-100 gl) is introduced by inserting
the tip of a micro-pipette through a port in the outer (water) jacket, and into the
gas inlet orifice in the centre of the graphite tube. The graphite cylinder is then
heated by the passage of an electric current to a temperature which is high enough
to evaporate the solvent from the solution. The current is then increased so that
firstly the sample is ashed, and then ultimately it is vapourised so that metal
atoms are produced, typically at a temperature of about 3000 K. For
reproducibility, the temperatures and the timing of the drying, ashing and
atomisation processes must be carefully selected according to the metal which is
to be determined. The absorption signals produced by this method may last for
several seconds and can be recorded on a chart recorder. Each graphite tube can
be used for 100-200 analyses depending upon the nature of the material to be
determined.

(ii) The graphite rod, A graphite rod of 2 mm diameter was introduced by
West (Ref. 4) as a means of producing atoms from the sample, and a commercial
device is now available from Messrs Shandon Southern of Camberley, Surrey.
The sample is placed upon the rod which is heated, typically by a current of 100 A
from a low voltage (5 V) supply. The rod is placed just below the path of the beam
from the z'adiation source so that vapour from the sample can move upwards into
the beam and its absorbance be measured. The whole assembly is contained in a
chamber fitted with quartz windows which is purged with argon.

In some circumstances it is found advantageous to coat graphite rods (or
tubes) with a layer of pyrolytic graphite: this leads to improved sensitivity with
elements such as vanadium and titanium which are prone to carbide formation.

The main advantages of flameless techniques is that very small samples (as low
as 0.5 /il) can be analysed, and very little or no sample preparation is needed: in
fact certain solid samples can be analysed without prior dissolution. It should
however be appreciated that greater expertise is required for flameless techniques,
and they should be regarded as complementary to the usual flame methods.

Amongst other devices used to produce the required atoms in the vapour state

818

FLAME SPECTROMETRY XXn,7

are the Delves Cup which enables the rapid determination of lead in blood
samples ; the sample is placed in a small nickel cup wliich is then inserted directly
into an acetylene-air flame. The tantalum boat is a similar device to the Delves
cup; in this case the sample is placed in a small tantalum dish which is then
inserted into an acetylene-air flame.

(b) Vapour technique. This procedure is strictly conflned to the
determination of mercury (Ref. 14), which in the elemental state has' an
appreciable vapour pressure at room temperature so that gaseous atoms exist
without the need for any special treatment. As a method for determining mercury
compounds, the procedure consists in the reduction of a solution of a mercury(II)
compound with tin(II) chloride to form elemental mercury. A diagram of a
suitable apparatus (the Rank Hilger H1469 Atomspek accessory) for the

determination of mercury is shown in

<Z3

Lamp

Absorption i

-dCD

Lamp

Reaction

vessel

Fig.xxn,7

Fig. XXII, 7.

This apparatus may also be adapted for
the determination of arsenic, antimony, and
selenium by conversion to their volatile
hydrides by the use of sodium borohydride
as reducing agent. In these cases the method
differs from that for the determination of
mercury since the hydrides thus' formed
cannot be examined directly in the absorp-
tion tube, but they are readily dissociated
I into atoms in an argon-hydrogen flame. The
requisite additional apparatus is indicated
by the dashed lines in Fig. XXII, 7.

XXn, 7. RESONANCE LINE SOURCES.
As indicated in Fig. XXII, 3, for both atomic
absorption spectroscopy and atomic
fluorescence spectroscopy a resonance line
source is required, and the most important of
these is the hollow cathode lamp which is
shown diagrammatically in Fig. XXII, 8. For
any given determination the hollow cathode
lamp used has an emitting cathode of the
same element as that being studied in the
flame. The cathode is in the form of a

cylinder, and the electrodes are enclosed in a
borosilicate or quartz envelope which contains an inert gas (neon or argon) at a
pressure of approximately 5 torr. The application of a high potential across the
electrodes causes a discharge which creates ions of the noble gas. These ions are
accelerated to the cathode and on collision, excite the cathode element to
eimssion. Multi-element lamps are available in which the cathodes are made from
alloys, but in these lamps the resonance line intensities of individual elements are
somewhat reduced.

Electrodeless discharge lamps were originally developed as radiation sources
mr atomic absorption spectroscopy and atomic fluorescence spectroscopy by
agnail et al. (Ref. 5) ; they give radiation intensities which are much greater than

819

XXII, 8/9 QUANTITATIVE INORGANIC ANALYSIS

Fig.XXn,8

those given by hollow cathode lamps. The electrodeless discharge lamp consists
of a quartz tube 2-7 cm in length and 8 mm in internal diameter, containing up to
20 mg of the required element or of a volatile salt of the element, commonly the
iodide; the tube also contains argon at low pressure (about 2 torr). Under
operating conditions the material placed in the tube must have a vapour pressure
of about 1 mm at a temperature of 200-400 °C. A microwave frequency of 2000
to 3000 MHz applied through a wave guide cavity provides the energy of
excitation.

XXII, 8. MONOCHROMATOR. The purpose of the monochromator is to
select a given emission line and to isolate it from other lines and occasionally,
from molecular band emissions.

In a simple flame (emission) photometer, e.g., the Corning-EEL Model 100
Flame Photometer (see Section XXII, 12), an interference filter (Section XVm, 7)
is used. In more sophisticated flame emission spectrophotometers which require
better isolation of the emitted frequency, a prism or a grating monochromator is
employed, and a resolution of 0.1 nm should be achieved.

In atomic absorption spectroscopy the function of the monochromator is to
isolate the resonance line from all non-absorbed lines emitted by the radiation
source. In most commercial instruments diffraction gratings (Section XVIII, 7)
are used because the dispersion provided by a grating is more uniform than that
given by prisms, and consequently grating instruments can maintain a higher
resolution over a longer range of wavelengths.

XXn, 9. DETECTORS. For the simple flame emission photometer (Section
XXn, 12) a barrier layer cell (Section XVIII, 7) is a sufficiently good detector
because an Intense wide band of energy reaches the detector. In atomic
absorption spectrophotometers, in view of the improved spectral sensitivity
required, photomultipliers (Section XVni, 7) are employed. The output from the

820

FLAME SPECTROMETRY XXH, 10

detector, is fed to a suitable read-out system, and in this connection it must be
borne in mind that the radiation received by the detector originates not only from
the resonance line which has been selected, but may also arise from emission
within the flame. This emission can be due to atomic emission arising from atoms
of the element under investigation, and . may also arise from molecular band
emissions. Hence instead of an absorption signal intensity J^, the detector may
receive a' signal of intensity (/^ -t- S) where S is the intensity of emitted radiation.
Since only the measurement arisi ng from the resonance line is required, it is
important that this be distinguished from the effects of flame emission. This is
achieved by modulation of the emission from the resonance line source by either a
mechanical chopper device, or electronically, by using an alternating current
signal appropriate to the particular frequency of the resonance line, , and the
detector amplifier is then tuned to this frequency : in this way, the signals arising
from the flame, which are essentially d.c. in character, are effectively removed.

The read-out systems available include meters, chart recorders, and digital
display ; meters have now been virtually superseded by the alternative methods of
data presentation.

XXn, 10. INTERFERENCES. Various factors may affect the flame emis-
sion of a given element and lead to interference with the determination of the
concentration of a given element. These factors may be broadly classified as (a)
spectral interferences and (b) chemical interferences.

Spectral interferences in AAS arise mainly from overlap between the
frequencies of a selected resonance line with lines emitted by some other element ;
this arises because in practice a chosen line has in fact a finite ‘band-width’. Since
in fact the line width of an absorption line is about 0.005 nm only a few cases of
spectral overlap between the emitted lines of a hollow cathode lamp and the
absorption lines of metal atoms in flames have been reported: Table XXII, 3
includes some typical examples of spectral interferences which have been
observed (see Refs. 6, 7, 8, 9). However most of this data relates to relatively minor
resonance lines and the only interferences which occur with preferred resonance
lines are with copper where europium at a concentration of about 150 p.p.m.
would interfere, and mercury where concentrations of cobalt higher than 200
p.p.m. would cause interference.

Table XXII, 3 Some typical spectral interferences

Resonance

source

Wavelength

nm)

Analyte

Wavelength
(X nm)

Aluminium

308.216

Vanadium

308.211

Antimony

,231.147

Nickel

231.095

Copper

324.754

Europium

324.755

Uallium

Iron

Mercury

403.307

Manganese

403.307

271.903

253.652

Platinum

Cobalt

271.904

253.649

With flame emission spectroscopy, there is greater likelihood of spectral
interferences when the line emission of the element to be determined and those
ue to interfering substances are of similar wavelength, than with atomic
a sorption spectroscopy. Obviously some of such interferences may be
e iminated by improved resolution of the instrument, e.g., by use of a prism

821

XXn, 11 QUANTITATIVE INORGANIC ANALYSIS

rather than a filter, but in certain cases it may be necessary to select other, non-
interfering lines for the determination. In some cases it may even be necessary to
separate the element to be determined from interfering elements by a separation
process such as ion exchange or solvent extraction (see Chapters VI, VII).

Apart from the interferences which may arise from other elements present in
the substance to be analysed, some interference may arise from the emission band
spectra produced by molecules or molecular fragments present in the flame
gases : in particular, band spectra due to hydroxyl and cyanogen radicals arise in
many flames. Although in AAS these flame signals are not modulated (Section
XXII, 9), in practice care should be taken to select an absorption line which does
not correspond with the wavelengths due to any molecular bands because of the
excessive ‘noise’ produced by the latter: this leads to decreased sensitivity and to
poor precision of analysis.

XXn, 11. CHEMICAL INTERFERENCES. The production of ground
state gaseous atoms which is the basis of flame spectroscopy may be inhibited by
two main forms of chemical interference: (a) by stable compound formation, or
(b) by ionisation.

(a) Stable compound formation leads to incomplete dissociation of the
substance to be analysed when placed in the flame, or it may arise from the
formation within the flame of refractory compounds which fail to dissociate into
the constituent atoms. Examples of these types of behaviour are shown by (i) the
determination of calcium in the presence of sulphate or phosphate, and (ii) the
formation of stable refractory oxides of titanium, vanadium, and aluminium.
Chemical interferences can usually be overcome in one of the following ways.

A. Increase in flame temperature often leads to the formation of free gaseous
atoms, and for example aluminium oxide is more readily dissociated in an
acetylene-nitrous oxide flame than it is in an acetylene-air flame. A
calcium-aluminium interference arising from the formation of calcium
aluminate can also be overcome by working at the higher temperature of an
acetylene-nitrous oxide flame.

B. By the use of ‘Releasing Agents’. If we consider the reaction

then it is clear that an excess of the releasing agent (R) vrill lead to an enhanced
concentration of the required gaseous metal atoms M : this will be especially true
if the product R — X is a stable compound. Thus in the determination of calcium
in the presence of phosphate, the addition of an excess of lanthanum chloride or
of strontium chloride to the test solution will lead to formation of lanthanum
(or strontium) phosphate, and the calcium can then be determined in an
acetylene-air flame without any interference due to phosphate. The addition of
EDTA to a calcium solution before analysis may increase the sensitivity of the
subsequent flame spectrophotometric determination : this is possibly due to the
formation of an EDTA complex of calcium which is readily dissociated in
the flame.

C. Extraction of the analyte or of the interfering element(s) is an obvious
method of overcoming the effect of ‘interferences’. It is frequently sufficient to
perform a simple solvent extraction to remove the major portion of an interfering
substance so that, at the concentration at which it then exists in the solution, the
interference becomes negligible. If necessary, repeated solvent extraction will
reduce the effect of the interference even further, and equally, a quantitative

822

FLAME SPECTROMETRY XXD, 11

solvent extraction procedure may be carried out so as to isolate the substance to
be determined from interfering substances. . • .

(b) Ionisation of the ground state gaseous atoms within a flame

will reduce the intensity of the emission of the atomic spectral hnes in flame
emission spectroscopy, or will reduce the extent of absorption in atomic
absorption spectroscopy. It is therefore clearly necessary to reduce the possibility
of ionisation occurring to a minimum, and an obvious precaution to take is to use
a flame operating at the lowest possible temperature which is satisfactory for the
element to be determined. Thus the high temperature of an acetylene-air or of an
acetylene-nitrous oxide flame may result in the appreciable ionisation of
elements such as the alkali metals and of calcium, strontium and barium. The
ionisation of the element to be determined may also be reduced by the addition of
an excess of an ionisation suppressant; this is usually a solution containing a
cation having a lower ionisation potential than that of the analyte. Thus, for
example, a solution containing potassium ions at a concentration of 2000 p.p.m.
added to a solution containing calcium, barium, or strontimn ions creates an
excess of electrons when the resulting solution is nebulised into the flame, and this
has the result that the ionisation of the metal to be determined is virtually
completely suppressed.

In addition to the compound formation and ionisation eSects which have been
considered, it is also necessary to take account of so-called Matrix effects. These
are predominantly physical factors which will influence the amount of sample
reaching the flame, and are related in particular to factors such- as the viscosity,
the density, the surface tension and the volatility of the solvent used to prepare
the test solution. If we wish to compare a series of solutions, e.g., a series of
standards to be compared with a test solution, it is clearly essential that the same
solvent be used for each, and the solutions should not differ too widely in their
bulk composition.

In some circumstances interference may result from molecular absorptions.
Thus, for example, in an acetylene-air flame a high concentration of sodium
chloride will absorb radiation at wavelengths in the neighbourhood of 213.9 nm
which is the wavelength of the zinc resonance line: hence sodium chloride would
represent an interference in the determination of zinc under these conditions.
Such interferences can usually be avoided by choosing a different resonance
wavelength for carrying out the determination, or alternatively by using a
different flame so that the operating temperature is increased thus leading to
dissociation of the interfering molecules.

To summarise, it may be stated that almost all interferences eneountered in
atomic absorption spectroscopy can be reduced, if not completely eliminated by
the following procedures.

1. Ensure if possible that standard and sample solutions are of similar bulk
composition to eliminate matrix effects.

2. /^teration of flame composition or of flame temperature can beused to reduce
the likelihood of stable compound formation within the flame.

j. Selection of an alternative resonance line will overcome spectral interferences
uom other atoms or molecules and from molecular fragments. . .

■ Occasionally, separation, e.g., by solvent extraction or by an ion exchange
process, may be necessary to remove an interfering element; such separations

823

XXn, 12 QUANTITATIVE INORGANIC ANALYSIS

are most frequently necessary when dealing with flame emission spectroscopy.

It may also be noted that the interference referred to as background absorption,
which arises from the presence in the flame of gaseous molecules, molecular
fragments, and in some instances of smoke, is dealt with in many modem
instruments by the incorporation of a background correction facility: a typical
example is discussed in Section XXII, 14.

With regard to the relative merits of the FAAS and FES procedures, it may be
stated in general terms that FAAS is a more selective technique than FES, and in
terms of sensitivity it is also to be preferred when we are dealing with lines of
wavelengths less than about 350 nm. However, for lines of wavelengths
appreciably greater than 350 nm, then FES is the more sensitive technique. These
general conclusions may be illustrated by the following data relating to some
typical metals;

zinc line 213.9 nm; sensitivity 0.009 (FAAS), 80 (FES);
magnesium line 285.2 nm ; sensitivity 0.003 (FAAS), 1 .0 (FES) ;
calcium line 422.7 nm ; sensitivity 0.02 (FAAS), 0.0 1 (FES) ;
sodium line 589.0-nm ; sensitivity 0.003 (FAAS), 0.001 (FES);
lithium line 670.8 nm ; sensitivity 0.02 (FAAS), 0.007 (FES).

Note however that since the flame conditions differ for each of the above
elements, it is not possible to make an absolute comparison of sensitivities; the
figures quoted do nevertheless serve as a rough guide for the comparison of
sensitivities.

Commercially available instruments

In the following Sections will be found brief descriptions of a selection of
commercially available instruments. The general mode of operation will be
apparent from the details given later in the Experimental Sections, but for any
particular instrument the handbook supplied by the manufacturer must be
consulted.

XXn, 12. FLAME PHOTOMETERS. A flame photometer can be com-
pared to a photoelectric absorptiometer and the intensity of the filtered
radiation from the flame is measured with a photoelectric detector. The filter,
interposed between the flame and the detector, transmits only a strong line of the
element. The simplest and least-expensive detector is a barrier-layer cell (Section
XVin, 7) : if suflJcient energy reaches the cell no amplification or external power
supply is necessary, and only a sensitive galvanometer is required. The barrier-
layer cell has a high temperature coefficient: it must therefore be placed at a
cool part of the photometer. In some cases the precision is improved by the use of
an internal standard and two filters and, in general, two photocells (one for the
standard and one for the unknown) are utilised ; the electronic circuit can be
devised to give a direct reading of the ratio of line intensities. Flame photometers
are intended primarily for analysis of sodium and potassium and also for calcium
and lithium, i.e., elements which have an easily excited flame spectrum of
sufiicient intensity for detection by a photocell. The lay-out of a simple flame
photometer is shown in Fig. XXII, 9. Air at a given pressure is passed into an
atomiser and the suction this produces draws a solution of the sample into the
atomiser, where it joins the air stream as a fine mist and passes into the burner.

824

FLAME SPECTROMETRY XXH, 12

Here, in a small mixing chamber, the air meets the fuel gas supplied to the burner
at a given pressure and the mixture is burnt. Radiation from the resulting flame
passes through a lens, then through an iris diaphragm, and finally through an
optical filter which permits only the radiation characteristic of the element under
investigation to pass through to the photocell. The output from the photocell is
measured on a suitable galvanometer. The flame is surrounded by a chimney to
protect it from draughts. The optical path from the chimney to the photocell is

enclosed in a light-tight box.

Fig.xxn,10

An example of an instrument of
this type is the Corning EEL
Model 100 Flame Photometer
which is depicted in Fig. XXII, 10;
a line diagram of the essential
parts is shown in Fig XXII, 1 1 . It is
a simple single-cell photometer
and will operate satisfactorily with
coal gas (or propane or butane or
‘Calor’ gas) and compressed air.
The elements which can be de-
termined are sodium, potassium,
calcium, and lithium. The regions
of the spectrum appropriate to the
elements being determined are
isolated by means of optical filters.

The operation of the instrument
will be understood by reference to
Fig. XXII, 11. Airis introduced to
the all-metal atomiser 1 through a
control valve 2 at a pressure
indicated on a gauge. 3 mounted
on the , front of the instrument.

825

XXn, 13 QUANTITATIVE INORGANIC ANALYSIS

Liquid from the beaker 4 containing the sample is drawn up the inlet tube 5 by
the stream of air which atomises the sample to a fine mist. The atomiser clips into
a plug 6 at one end of the spray chamber 7; in which the larger droplets fall from
the air stream and flow to waste through the drain tube 8. Gas is introduced into ,
the spray chamber through the inlet tube 9, which is connected by tubing to the
automatic gas pressure stabiliser 10 and control valve 11. The gas/air mixture
burns in a broad flat flame, and hot gases pass up a well-ventilated chimney 12.
The light emitted by the flame is collected by a reflector 13 and focused by a lens
14 through the interchangeable optical filters 15 on to a barrier-layer photocell
16. The current generated by this photocell is taken through a calibrated
potentiometer 17 to a suspension galvanometer unit 18. A glass window 19 is
interposed between the lens and the filter for cooling purposes. A cover plate 20
having a V-shaped recess is provided to locate the 10-cm^ beaker holding the
sample solution, which may be slid up and held against the stop when the sample
is to be sprayed.

4 5 167 9 16151914 1213

Fig.xxn,!!

For further details of the instrument see Section XXII, 21.

XXn. 13. SINGLE BEAM ATOMIC ABSORPTION SPECTROPHOTO-
METERS. Many commercial instruments are based on the use of a single beam,
modulated a.c. system, and a typical example of such an instrument is the Hilger
and Watts Atomspek H1550, a line diagram of which is shown in Fig. XXII, 12.
The important features of this instrument include an easily accessible sample
area which can house an automatic sampler capable of handling 40 samples and 8
reference checks in six minutes; the sample volume required is only 0.3 cm^. A
turret holding six hollow cathode lamps is provided with an independent current-
stabilised supply to each lamp. The instrument is capable of resolving lines which
are less than 0.1 nm apart and is equipped with a photo-multiplier that functions
well over the wave-length range of 193 to 853 nm. Integrated measurements can
be made with high precision by making use of the crystal-operated digital clock

826

FLAME SPECTROMETRY XXH, 14

which is included; this clock controls the stabilised wave forms for pulsing the
hollow cathode lamp supplies as well as the signal demodulation. The analytical
data obtained can be recorded on a chart, or printed automatically by digital
typewriter.

Particular features of this instrument are that it can also be used as a flame
emission spectrophotometer, and that accessories are available for flameless
techniques including the vapour technique for mercury, arsenic and tellurium
(see Section XXH, 6).

The Varian Techtron Model AA-6 single beam atomic absorption
spectrophotometer is shown in Fig. XXII, 13. A feature of this instrument is the
incorporation of an ‘optical rail’ on to which many of the optical and sampling
components are fixed so that it is a simple matter to rearrange these to suit
specific requirements; a turret holding four hollow cathode lamps is also
attached to this rail. The burners are made of titanium and are designed to handle
solutions which are of high solids content without clogging of the burners. The
monochromator provides uniform dispersion over the wavelength range
185-1000 nm. The presentation of data may be by means of a meter, by a digital
display or by means of a digital printer, and facilities are provided for integrating
three alternative time periods (3, 10 or 30 seconds).

This instrument can also be used as a flame emission spectrophotometer, and
permits the use of graphite rod techniques.

14. DOUBLE BEAM ATOMIC ABSORPTION SPECTRO-
With the introduction of a relatively ‘low noise’ burner in
o2 having a constant aspiration rate, it was no longer the burner which was the
mam source of instrument instability ; the limiting factor was now the stability of
e ollow cathode lamp. The double beam system was designed to overcome the
Elm intensity. A typical instrument of this type is the Perkin

er Model 460 atomic absorption spectrophotometer, a simplified optical

827

82 ;

FLAME SPECTROMETRY XXII, 14

diagram of which is showa in Fig. XXII, 14, and the instrument itself in Fig.
XXII, 15. In this instrument, a rotating chopper passes the beam from the lamp
alternately through the flame (to give the sample beam), and around the flame (to
give the reference beam). The sample and reference beams are re-combined by a
half silvered mirror, and are then passed together to the photomultiplier
detector. The electrieal circuits are designed to measure the ratio of the two
beams, and hence variations iii lamp intensity, photomultiplier sensitivity and
electronic gain affects both signals similarly, and the ratio of the two signals
compensates for variation in the quantities listed.

Photomultiplier

D2 Arc

1 1

Primary source

Fig.xxn, 14

h ^ 1 ' '.Z£

Fig-Xxn,15

e burner chamber of the Model 460 is made of stainless steel lined with a
corrosion-resistant plastic; it is provided with a spring-loaded check valve which
pens automatically in the rare event of an explosive flash-back and protects the
burner is so designed that only the smallest droplets
the flame, thus reducing background absorption and providing optimum

829

xxn, 15/16 QUANTITATIVE INORGANIC ANALYSIS

precision. The uptake to the nebuliser can be varied continuously thus enabling
low sample consumption when the quantity available is limited. Provision is
made for the connection of one fuel gas and two oxidant gases with rapid
switching from one to the other, and an automatic check that the correct burner
head is in position before nitrous oxide can be switched on ; the burner heads are
constructed of titanium.

A background corrector is incorporated which takes the form of a high
intensity deuterium arc lamp, producing an emission continuum which travels
the same double beam path as does the light from the resonance source (see Fig.
XXII, 14). The background absorption affects both the sample and reference
beams and so when the ratio of the intensities of the two beams is taken, the
background effects are eliminated.

A special feature of the Model 460 is the inclusion of a microcomputer, and
this makes it a very simple matter to calibrate the instrument to read directly in
concentration units. The concentration values of standard solutions are fed to
the computer through the numerical keyboard; normally three standard
solutions may be used, but under some conditions a single standard will suffice. A
solution blank aspirated through the burner automatically adjusts the zero of the
instrument, and on aspiration of the standards, a working curve is automatically
computed.

xxn, 15. ATOMIC FLUORESCENCE SPECTROSCOPY. Within the
confines of the present volume it is not possible to provide a detailed discussion of
instrumentation for atomic fluorescence spectroscopy. An instrument for
simultaneous multi-element determination described by Mitchell and Johansson
(Ref. 10) has been developed commercially. Many atomic absorption
spectrophotometers can be adapted for fluorescence measurements and details
are available from the manufacturers. Detailed descriptions of atomic
fluorescence spectroscopy are to be found in many of the volumes listed in the
Bibliography (Section XXII, 28).

Experimental

xxn, 16. EVALUATION METHODS. Before dealing with the experimen-
tal details of AAS or FES determinations it is necessary to consider the mode of
treatment of the experimental data obtained. To convert the measured
absorption values into the concentration of the substance being determined it is
necessary either to make use of a calibration curve, or to carry out the ‘standard
addition’ procedure.

(a) Calibration curve procedure. A calibration curve for use in atomic
absorption or in flame emission measurements is plotted by aspirating into the
flame samples of solutions containing known concentrations of the element to be
determined, measuring the absorption (emission) of each solution, and then
constructing a graph in which the measured absorption (emission) is plotted
against the concentration of the solutions. If we are dealing with a test solution
which contains a single component then the standard solutions are prepared by
dissolving a weighed quantity of a salt of the element to be determined in a known
volume of distilled (de-ionised) water in a graduated flask. If however other
substances are present in the test solution, then these should also be incorporated

830

FLAME SPECTROMETRY XXH, 17

in the standard solutions and at a similar concentration to that existing in the test
solution. At least four standard solutions should be used covering the optimum
absorbance range of 0.2 to 0.8, and if the calibration curve is found to be non-
linear (this often happens at high absorbance values), then measurements with
additional standard solutions should be carried out. In common with all
absorbance measurements, the readings must be taken after the instrument zero
has been adjusted against a ‘blank’ which may be either distilled water, or a
solution of similar composition to the test solution but minus the component to
be determined.- It is usual to examine the standard solutions in order of increasing
concentration, and after making the measurements with one solution, distilled
water is aspirated into the, flame to remove all traces of solution before
proceeding to the next solution. At least two, and preferably three, separate
absorption readings should be made with each solution, and an average value
taken.

If necessary, the test solution must be suitably diluted using a pipette and a
graduated flask, so that it too gives absorbance readings in the range 0.2-0.8.
Using the calibration curve it is a simple matter to interpolate from the measured
absorbance of the test solution the concentration of the relevant element in the
solution. The working graph should be checked occasionally by making
measurements with the standard solutions, and if necessary a new calibration
curve must be drawn.

{b) The standard addition technique. When dealing with a test solution
which is complex in character, or one whose exact composition is unknown, it
may be very difficult and even impossible to prepare standard solutions having a
similar composition to the sample. In such a case the method of standard
addition can be employed. As described in Section XVI, 5 , this involves the
addition of known amounts of the ion to be determined to a number of aliquots
of the sample solution ; the solutions thus obtained should all be diluted to the
same final volume. Naturally, if the absorbance of the test solution is too high, a
quantitative dilution must be carried out, and the measurements made with this
diluted solution. The absorbance of the test solution is first measured, and then
each of the prepared solutions is examined in turn, leading up to the solution of
highest concentration, and remembering to aspirate distilled water into the flame
between each solution. The absorbance values are then plotted against the added
concentration values ; a straight line plot should result and the straight line can be
extrapolated to the concentration axis — the point where the axis is cut gives the
concentration of the test solution. If the graph is non-linear, then extrapolation
cannot be undertaken with any confidence and it is important to realise that an
extrapolation procedure is never as reliable as interpolation, and the latter
should therefore be chosen if at all possible.

xxn, 17. preparation of sample solutions. For the appli-
cation of flame spectroscopic methods the^aipple must be prepared in the form
of a suitable solution unless it is already pre^ted in this formTexCeptionalljr,
solid samples can be handled directly in some of the non-flame techniques

(Section xxn, 6).

Aqueous solutions may sometimes be analysed directly without any pre-
treatment, but it is a matter of chance that the given solution should contain
the correct amount of material to give a satisfactory absorbance reading. If the
existing concentration of the element to be determined is too high then the

831

XXU, 18 QUANTITATIVE INORGANIC ANALYSIS

solution must be diluted quantitatively before commencing the absorption
measurements. Conversely, if the concentration of the metal in the test solution is
too low, then a concentration procedure must be carried out (see below, under
Separatory Methods).

Solutions in organic solvents may, with certain reservations, be used directly
provided that the viscosity of the solution is not very different from that of an
aqueous solution. The important consideration is that the solvent should not
lead to any disturbance of the flame; an extreme example of this is carbon
tetrachloride which may extinguish an air-acetylene flame. In many cases,
suitable organic solvents {e.g., 4-methylpentan-2-one (methyl isobutyl ketone)
and the hydrocarbon mixture sold as ‘white spirit’} give enhanced production of
grormd state gaseous atoms and lead to about three times the sensitivity which is
achieved with aqueous solutions. Due regard must of course be paid to the
question of safety: see Section XXII, 19.

Inorganic solids such as metallic alloys, minerals, cements, etc., must be
brought into solution by the usual standard techniques, the aim being to produce
a clear solution with no loss of the element to be determined. Generally speaking,
the final solution should not contain acid at a greater concentration than about
\M since the aspiration of extremely corrosive solutions into the burner of the
apparatus should be avoided as far as possible: the instruction manual supplied
with the instrument will normally give guidance in this direction.

Organic solids which contain trace elements can sometimes be dissolved in a
suitable organic solvent, or alternatively the organic material may be oxidised
and the residue treated to give an aqueous solution of the element to be
determined.

Separation techniques may have to be applied if the given sample contains
substances which act as interferences (Section XXU, 10), or, as explained above,
if the concentration of the element to be determined in the test solution is too low
to give satisfactory absorbance readings. As already indicated (Section XXII,
10 ), the separation methods most commonly used in conjunction with flame
spectrophotometric methods are solvent extraction (see Chapter VI) and ion
exchange (Chapter VII). When a solvent extraction method is used, it may
happen that the element to be determined is extracted into an organic solvent,
and as discussed above it may be possible to use this solution directly for the
flame photometric measurement.

XXn, 1& PREPARATION OF STANDARD SOLUTIONS. In flame
spectrophotometric measurements we are concerned with solutions having very
small concentrations of the element to be determined. It follows that the standard
solutions which will be required for the analyses must also contain very small
concentrations of the relevant elements, and it is rarely practicable to prepare the
standard solutions by weighing out directly the required reference substance. The
usual practice therefore is to prepare stock solutions which contain about
1000/igcm~^ of the required element, and then the working standard solutions
are prepared by suitable dilution of the stock solutions. Solutions which contain
less than 10//gcm~^ are often found to deteriorate on standing owing to
adsorption of the solute on to the walls of glass vessels. Consequently, standard
solutions in which the solute concentration is of this order should not be stored
for more than 1 to 2 days.

The stock solutions are ideally prepared from the pure metal or from the pure

832

FLAME SPECTROMETRY XXH, 19

metal oxide by dissolution in a suitable acid solution; the solids used must of
course be of the highest purity, e.g., the Johnson Matthey ‘Specpure’ raiige of
reagents. In many cases however it is prepared, by dissolution of a suitable
metallic salt in de-ionised water provided that the salt satisfies the normal
requirements of a primary standard.

XXn, 19. SAFETY PRACTICES. Before commencing any experimental
work with either a flame (emission) photometer or an atomic absorption
spectrophotometer, the following guide lines on safety practices should be
studied. These recommendations are a summary of the Code of Practice
recommended by the Scientific Apparatus Makers Association (SAMA) of the
USA; for full details see Ref. 11.

1 . Ensure that the laboratory in which the apparatus is housed is well ventilated
and is provided with an adequate exhaust system having air-tight joints on
the discharge side; some organic solvents, especially those containing
chlorine, give toxic products in a flame.

2. Gas cylinders must be fastened securely in an adequately ventilated room
well away from any heat or ignition sources. The cylinders must be clearly
marked so that the contents can be immediately identified.

3. When the equipment is turned off, close the fuel gas cylinder valve tightly and
bleed the gas line to the atmosphere via the exhaust system.

4. The piping which carries the gases from the cylinders must be securely fixed
in such a position that it is unlikely to sulfer damage.

5. Make periodic checks for leaks by applying soap solution to joints and seals.

6. The following special precautions should be observed with acetylene.

(a) Never run acetylene at a pressure higher than ISp.s.i. (103 kN m~^); at
higher pressures acetylene can explode spontaneously.

(h) Avoid the use of copper tubing. Use tubing made from brass containing
less than 65 per cent copper, from galvanized iron or from any other
material that does not react with acetylene.

(d) Never run an acetylene cylinder after the pressure has dropped to 50 p.s.i.
(3430 kN m~^); at lower pressures the gas will be contaminated with
acetone.

8 .

10.

A nitrous oxide cylinder should not be used after the regulator gauge has
dropped to a reading of 100 p.s.i. (6860 kN m~^).

A burner which utilises a mixture of fuel and oxidant gases and which is
attached to a waste vessel (liquid trap) should be provided with a U-shaped
connection between the trap and the burner chamber. The head of liquid in
the connecting tube should be greater than the operating pressure of the
burner: if this is not achieved, mixtures of fuel and oxidant gas may be
vented to the atmosphere and form an explosive mixture. The trap should be
made of a material that will not shatter in the event of an explosive flash-back
in the burner chamber.

Care must be exercised when using volatile inflammable organic solvents for
aspiration into the flame. A container fitted with a cover which is provided
with a small hole for the sample capillary is recommended.

Never view the flame or hollow cathode lamps directly; protective eye wear
should always be worn. Safety spectacles will usually provide adequate

833

XXn; 20 QUANTITATIVE INORGANIC ANALYSIS

protection from ultraviolet light, and will also provide protection for the eyes
in the event of the apparatus being shattered by an explosion,
n. ' Never leave aflame unattended. ■ . • '

Some selected determinations

XXII, 20. INTRODUCTION. It is impossible in the present volume for the
determination of a wide range of elements by atomic absorption spectroscopy to
be discussed in detail. A few detailed examples of the application of atomic
absorption and atomic emission methods are given in Sections XXn, 21-26;
these have been chosen to illustrate the general procedures involved, including
the manner in which certain interferences may be overcome and how chemical
pre^treatraent is often necessary in order to perform a successful analysis by this
technique.

In Table XXII, 4 is listed the wavelength of the most widely used resonance line
for all the common elements, together with the normal composition of the flame
gases. The optimum working range of concentrations is quoted, and although
this can vary with the instrument used, the values cited may be regarded as
typical. The term sensitivity in atomic absorption spectroscopy is defined as the
concentration of an aqueous solution of the element which absorbs 1 per cent of
the incident resonance radiation; in other words, it is the concentration which
gives an absorbance of 0.0044. As a rough guide, the sensitivity may be taken as
about one-fiftieth of the lower value of the optimum absorbance range (Section
XXn, 16(a)). It should be noted that sensitivity is largely dependent upon the
reactions occurring in the flame, and is not strictly a characteristic of a given
instrument.

The detection limit is another value which is often quoted, and this may be
defined in a variety of ways. The most widely accepted definition is that the
detection limit is the smallest concentration of a solution of an element that can
be detected with 95 per cent certainty. This is the quantity of the element that
gives a reading equal to twice the standard deviation of a series of at least ten
determinations taken with solutions of concentrations which are close to thelevel
of the blank.

TableXXn, 4 FAAS data for the common elements

Element Wavelength of main Flame*

resonance line (2 nm)

Working range
|igcm"^

328.1

AA(L)

1-5

309.3

NA(R)

193.7

AH(R)'

249.8

NA(R)

400-600

553.6

NA(R)

10-40

234.9

NA{R)

1-5

AA(L)

422.7

NA(R)

1-4

228.8

AA{L)

AA(L)

3-12

357.9

AA(R)

2-8

AP(L)

324.7

AA(L)

2-8

834

FLAME SPECTROMETRY XXII, 21

Element

Wavelength of main Flame*

resonance line (A nm)

Working range
/igcm:^

248.3 .

AA(L)

2.5-10

294.4

AA(L)

50-200

265.2

NA(R) ■

70-280

253.7

■AA(L)

100-400

303.9

AA(L)

• 15-60

208.9

AA(R)

40-160

766.5

AP(L)

0,5-2

670.8

AP(L)

1-4

285.2

AA(L)

0. 1-0.4

279.5

AA(L)

1-4

313.3

NA(R)

15-60

589.0

AP(L)

.0.15-0.60

232.0

AA(L)

3-12

290.9

NA(R)

50-200

217.0

AA(L)

5-20

244.8

AA(L)

4-16

265.9

AA(L)

50-200

780.0

AP(L)

2-10

343.5

AA(L)

5-25

349.9

AA(L)

30-120

217.6

AA(L)

10-40

391.2

NA(R)

15-60

196.0

AH(R)

20-90

251.6

NA(R)

70-280

224.6

AH(R)‘

15-60

460.7

NA(L)

2-10

214.3

AA(L)

10-40

364.3

NA(R)

60-240

276.8

AA(L)

10-50

318.5

NA(R)

40-120

255.1

NA(R)

250-1000

410.2

NA(R)

200-800

213.9

AA(L)

0.4-1 .6

* Key. L = Fuel lean R = Fuel rich . . . ,

AA = Air/Acetylene NA = Nitrous oxide/ Acetylene '

AP = Air/Propane AH = Air/Hydrogen

Notes. ‘ If there are many interferences then NA is to be preferred.

^ The use of the non-flame mercury cell (Section lOfll, 6) is far more sensitive for the
determination of mercury. , , . >

The data presented in Table XXII, 4, in conjunction with the experimental
details given in Sections XXII, 21-26, will enable the determinatioa of most
elements to be carried out successfully,' For detailed accounts of the
determination of individual elements by atomic absorption spectroscopy, the
Bibliography (Section XXQ, 28) should be consulted. Tn addition, most
instrument manufacturers supply applications handbooks relative to their
apparatus in which full experimental details are given.'

XXn, 21. EXPERIMENTS WITH A SIMPLE FLAME PHOTOMETER.
The following account refers to the use of the Coming EEL Model 100 Flame
Photometer (see Section XXn, 12). Before attempting to rise the instruihent read

835

XXn, 21 QUANTITATIVE INORGANIC ANALYSIS

the instruction manual and Sections XXII, 12 and XXn, 19; Fig. XXII, 1 1 should
be consulted in conjunction with the operating instructions which are given.

1. Adjust the sensitivity control (1 7) to the minimum value. ■

2. Turn the gas supply on fully and light the gas at the burner with a lighted
taper.

3. Adjust the air supply from a cylinder of compressed air (itself fitted with the
usual gauges and controls) until the pressure indicated on the pressure gauge
mounted on the front of the instrument (3) attains a value of lOlb/sqin.
(690 kNm-^).

4. Charge the small sample beaker (4) with de-ionised water and place it in
position in the instruinent. The liquid is drawn up the inlet tube (5) by the
stream of air and is atomised to a fine mist.

5. Regulate the gas supply so that the blue cone of the flame just forms ten
separate cones, one to each burner hole.

6. Place the appropriate filter in position.

7. Aspirate a standard solution containing the ion to be determined and, by
means of the calibrated potentiometer (17), adjust the galvanometer spot to
read approximately full-scale deflection.

8. Aspirate de-ionised water and adjust the galvanometer spot to read zero by
means of the zero control (22).

9. Aspirate the standard solution again and readjust the sensitivity control (17)
for full-scale deflection of the galvanometer.

10. Check the zero by aspirating de-ionised water.

1 1 . Aspirate solutions of known concentration but less than that of the standard
solution, and note the galvanometer reading at each concentration. Plot the
galvanometer readings (abscissae) against the concentration (ordinates)
expressed, say, as p.p.m., and thus prepare a calibration curve for each
element. (It is advisable to measure the standard solution periodically in
order to check, and if necessary adjust, the full-scale deflection of the
galvanometer.)

12. Aspirate the unknown solution in the flame, note the galvanometer
deflection, and evaluate the concentration from the calibration curve.

If the metal content of the solution is completely unknown the approximate
concentration may be determined with the aid of the potentiometer sensitivity
control (17). Aspirate a standard solution of suitable strength into the flame to
obtain a reading on the scale at full sensitivity (potentiometer reading of unity).
Note the exact scale reading. T urn the sensitivity control right down ; aspirate the
unknown solution and gently turn the sensitivity control until exactly the same
galvanometer reading is obtained. Note the potentiometer reading; this gives
approximately the concentration of the unknown solution relative to the
standard solution. The unknown solution may then be diluted to give a reading
on the scale at full sensitivity ; the exact concentration may then be deduced from
the calibration curve. The final measurements should always be made by
comparing readings obtained on the galvanometer scale with the unknown and
standard solutions at the same potentiometer setting.

If it is known that the test solution contains a sufficient concentration of an
interfering substance to affect the reading it will be necessary to employ standard
solutions which also contain approximately the same concentration of the
interfering substance as is present in the sample. The ideal method of removing
interferences is to separate the element being determined by chemical means (e.g.,

836

FLAME SPECTROMETRY XXn,22

calcium as the oxalate), but this procedure is not always practicable.

Preparation of standard solutioiK for calibration curves. The following
concentrations are suitable : ^ .

(a) Sodium. Dissolve 2.542 g A.R. sodium chloride in 1 dm" de-ionised water

in a graduated flask. This solution contains the equivalent of 1.000 mg Na per.
cm". Dilute this stock solution to give four solutions containing 10, 5, 2.5, and
1 p.p.m. of sodium ions. ' . . ^ • j

(b) Potassium. Dissolve 1.909 g A.R. potassium chloride in 1 dm de-iomsed
water. This solution contains the equivalent of 1.000 mgK per cm". Dilute this
stock solution to give four solutions containing 20, 10, 5, and 2 p.p.m. of
potassium ions.

(c) Calcium. Dissolve 2.497 g A.R. calcium carbonate in a- little dilute
hydrochloric acid, and dilute to 1 dm" with de-ionised water. This stock solution
contains the equivalent of 1.000 mg Ca per cm". Dilute this solution to give
solutions containing 100, 50, 25, and 10 p.p.m. of calcium ions.

(d) Lithium. Dissolve 5.324 g pure lithium carbonate in a little dilute
hydrochloric acid and dilute to 1 dm" with de-ionised water. This solution
contains 1.000 mg Li per cm". Dilute the stock solution to give solutions
containing 20, 10, 5, and 2 p.p.m. of lithium ions.

Prepare calibration curves for each of the above four elements. With the aid of
these calibration curves, carry out the following simple determinations. ■ '

(1) Potassium in A.R. potassium sulphate. Weigh out accurately about 0.20 g
A.R. potassium sulphate and dissolve it in 1 dm" de-ionised water. Dilute
10.0 cm" of this solution to 100 cm", and determine the potassium with the flame
photometer using the potassium filter.

(2) Potassium and sodium in admixture. Mix suitable volumes of the above
stock solutions so that the resulting solution contains, say, 4-10 p.p.m. Na and
10-15 p.p.m. K. Determine the Na and K with the aid of the appropriate filters.
Compare the results obtained with the true values.

(3) Sodium, potassium, and calcium in admixture. Mix appropriate volumes of

the above stock solutions so that the test solution contains, say, 5 p.p.m. Na,
10 p.p.m. K, and 40 p.p.m. Ca. Determine the Na, K, and Ca with the aid of the
appopriate filters. Compare the results obtained with the true values. :

(4) Calcium in calcium carbonate. Determine the calcium in an analysed
sample of dolomite. Dissolve about 0.38 g, accurately weighed, in Kl-
hydrochloric acid, warm gently, filter through a quantitative filter paper, wash,
dilute the combined filtrate and washings to 1 dm". Measure the calcium content
of the resulting solution: use a calcium filter. Compare the value for Ca thus
obtained with the known Ca content.

xxn, 22. DETERMINATION OF MAGNESIUM AND CALCIUM IN TAP
WATER (AAS). The determination of magnesium in potable water is very
straightforward ; very few interferences are encountered when ■ using an
acetylene-air flame. The determination of calcium is however more complicated ;
many chemical interferences are encountered in the acetylene-air flame and the
use of ‘releasing agents’ such as strontium chloride, lanthanuni chloride, or
EDTA is necessary. Using the hotter acetylene-nitrous oxide flame the only
significant interference arises from the ionisation of calcium, and under these
conditions an ‘ionisation buffer’ such as potassium chloride is added to the test
solutions.

837

xxn, 22 QUANTITATIVE INORGANIC ANALYSIS

(a) Determination of magnesinm. Preparation of the standard solutions. A
magnesium stock solution (1000 mg dm is prepared by dissolving' 1.000 g
magnesium metal (A.R.) in 50 cm^ of 5M hydrochloric acid. After dissolution of
the metal the solution is transferred to a 1 dm^ graduated flask and made up to
the mark with distilled water. An intermediate stock solution containing
50mgMg^'^ dm“^ is prepared by pipetting 50 cm^ of the stock solution into a
1 dm^ graduated flask and diluting to the mark. Dilute accurately four portions
of this solution to give four standard solutions of magnesium, with known
magnesium concentrations lying within the optimum working range of the
instrument to be used (typically 0.1-0.4pg Mg^"^ cm"^).

Procedure. Although the precise mode of operation may vary according to
the particular instrument used, the following procedure may be regarded as
typical. Place a magnesium hollow cathode lamp in the operating position, adjust
the current to the recommended value (usually 2-3 mA), and select the
magnesium line at 285.2 nm using the appropri a^te mono chromator slit width.

Connect the appropriate gas supplies-to the'buriierTolfowing the instructions
detailed for the instrument, and adjust the operating conditions to give a fuel-
lean acetylene-air flame.

Starting with the least concentrated solution, aspirate in turn the standard
'magnesium solutions into the flame, and for each take three readings of the
absorbance; between each solution, remember to aspirate de-ionised water into
the burner. Finally read the abso'fbance of the sample of tap water; this will
usually require considerable dilution in order to give an absorbance reading lying
within the range of values recorded for the standard., solutions. Plot the
calibration curve and use this to determine the magnesium concentration of the
tap water.

If the magnesium content of the water is greater than 5pgcm“^ it might be
considered preferable to work with the less sensitive magnesium line at
wavelength 202.5 nm.

(b) Determination of calcium. Two procedures are described, (i) involving

the use of releasing agents, and (ii) involving the use of an ‘ionisation buffer’ ; the
latter is the preferred technique provided that an acetylene-nitrous oxide flame is
available. . ,

Preparation of the standard solutions. For procedure (i) it is necessary to
incorporate a releasing agent in the standard solutions. Three different releasing
agents may be used for calcium, (a) lanthanum chloride, {b) strontium chloride
and (c) EDTA; of these (a) is the preferred reagent, but {b) or (c) make
satisfactory alternatives.

(а) Prepare a lanthanum stock solution (50000 mg dm"^) by dissolving 67 g
of lanthanum chloride (LaClj THjO) in 100 cm^ of lAf nitric acid. Warm gently
to dissolve the salt, then cool the solution and make up to 500 cm^ in a graduated
flask.

( б ) A strontium stock solution is prepared by dissolving 76 g of A.R.
strontium chloride (SrCl 2 6 H 2 O) in 250 cm^ of de-ionised water and then making
up to 500 cm^ in a graduated flask.

(c) An EDTA stock solution is prepared by dissolving 75 g of EDTA
disodium salt (A.R. quality) in 800 cm^ of de-ionised water. Warm gently until
the salt is dissolved, then cool and make up to 1 dm^ in a graduated flask.

For procedure (ii) an ionisation buffer is required and this involves preparing a
potassium stock solution (lOOOOmgdm"^). Dissolve 9.6g of A.R. potassium

838

FLAME SPECTROMETRY XXn,22

chloride in de-ionised water and make up to 500 cm? in a graduated flask. ; ■ -
Prepare a calcium stock solution (1000mgdm~^) by dissolving 2.497 g of
dried a!R. calcium carbonate in a minimum volume of IM hydrochloric acid :
about 50 cm^ will be required. When dissolution is complete, transfer the
solution to a I dm^ graduated flask and make up to the mark with de-ionised;
water. An intermediate calcium stock solution is prepared by pipetting 50 cm^ of
the stock solution into a I dm^ flask and making up to the mark with de-ionised
water. ■ '

The working standard solutions . for procedure (i) contain between
IpgCa^'^cm”^ to 5;igCa^^ cm“^ and are prepared by mixing appropriate
volumes of the intermediate stock solution (measured with a grade A pipette),
with suitable volumes of the chosen releasing agent solution, and then rhaking up
to 50 cm^ in a graduated flask; the releasing agent solution is measured in a
25 cm^ measuring cylinder. Five standard solutions are prepared containing
respectively 1.0, 2.0, 3.0, 4.0, and 5.0 cm^ of the intermediate stock solution and
10 cm^ of releasing agent (a) or 5 cm^ of either reagent (b) or (c). A blank solution
is similarly prepared but without the addition of any of the intermediate calciuni
stock solution.

For procedure (ii) the working standard solutions are prepared as detailed for
procedure (i) except that the releasing agent solution is replaced by lOcm^ of the
stock potassium solution.

The unknown calcium solution (the tap water), will normally require to be
diluted in order that its absorbance reading shall lie on the calibration curve, and
the same amount of releasing agent {procedure (i)}, or of ionisation buffer
{procedure (ii)}, must be added as in the standard solutions. So, for example, if
the tap water contains about lOOpgcm'^ of calcium, 25 cm^ of it are pipetted
into a lOOcm^ graduated flask and rahde up to the majrk with de-ionised water.’
Then 5cm^ of this solution is pipetted into a 50 cm^ graduated flask, and- if
procedure (i) is being followed, lOcm^ of reagent (a) is added, or 5 cm^ of either
reagent (6) or (c) and then the solution is made up to the mark. If procedure (ii) is
being followed, then lOcm^ of the stock potassium solution are used in place of
the releasing agent. If any cloudiness should develop during the preparation of
the final solution, add 1 cm^ of \M hydrochloric acid before making up to the
mark.

Procedure (i). Set up a calcium hollow cathode lamp selecting the
resonance line of waveleng th 422.7 nm, and a fuel-lean acetylene-air flame
following the details given in the instrument manual. The calibration procedure
IS similar to that described above for magnesium, but the aspiration of de-ionised
water into the burnei^after-taklng the readings for.eacE solution is even more
important in thiscase owing to the relatively high concentrations of salts present
as releasing agent; remember that de-ionised water should be aspirated into the
burner for a few minutes at the conclusion of the series of readings.

Procedure (ii). Make certain that the instrument is fitted with the correct
burner for an acetylene-nitrous oxide flanie, then set the instrument up with the
calcium hollow cathode lamp, select the resonance line of wavelength 422.7 nm ,
controls as specified in the instrument manual to give a fuel-
nch flame. Take measurements with the blank, the standard solutions, and with
e test solution, all.of which contain the ‘ionisation buffer’ ; the need, mentioned
un er procedure (i) for adequate treatment with de-ionised water after each
measurement applies with equal force in this case. Plot the calibration graph and

XXn, 23/24 QUANTITATIVE INORGANIC ANALYSIS

ascertain the concentration of the unknown'solution. ■

XXn, 23^.-DETERMINATION OF VANADIUM IN LUBRICATING OIL

I XAAS). ^he oil is dissolved in white fflrit and -the-absorption.whieh this solution
gives rise to is~conipaieci with that produced from standards made up from
vanadium naphthenate dissolved in white spirit.

Preparation of the standard solutions. The standard solutions are prepared
from a solution of vanadium naphthenate in white spirit (Nuodex Ltd) which
contains about 3 per cent of vanadium. Weigh out accurately about 0.6 g of the
vanadium naphthenate into a lOOcm^ graduated flask and make up to the mark
with white spirit: this stock solution contains about 180;/gcm~^ of vanadium.
Dilute portions of this stock solution measured with the aid of a Grade A 50 cm^
burette to obtain a series of working standards containing from 10-40 pgcm~^
of vanadium.

Procedure. Weigh out accurately about-S-g-of-the-oil-sample, dissolve in a
small volume of whimj pirit a nd transfer to a 50 cm^ graduated flask; use the
same solventTo' wash out the weighing bottle and finally to make up the solution
to the mark.

A doi,ihle-he am atomic absorption spectrophotometer should. be us ed, e.g.,
Perkin filmer Model 306 or Model 460 or equivalent instrument. Set up a
vana dium hollow cathode lamp selecting the resonance line of wavelen gth
318.^m, and adjust the gas controls to given fuel-ric h qpjylene^mitrQii.s-oxide
JRSelin accordance with the instruction manual. Aspirate successively into the
flame the solvent blank, the standard solutions, and finally the test solution, in
each case recording the absmSancer- reading. Plot the wTibration curve and
ascertain the vanadium content of the oil.

XXn, 24. determination of trace lead IN A FERROUS ALLOY
(AAS). The procedure followed entails the removal of gross interferents by
solvent extraction, and the selective extraction and concentration of the trace
metal by use of a chelating agent. The alloy used should not contain more than
0.1 g of copper in the sample weighed out.

Preparation of solutions. The following solutions are required.

Ammonia solution (concentrated, ‘0.880’, about 35 per cent NH3) either A.R.
or preferably the special atomic absorption spectroscopy (A.A.S.) reagent
obtainable from laboratory supply houses.

Hydrochloric acid, concentrated, A.R. or preferably A.A.S. reagent ; and also a
solution prepared by measuring 50 cm^ (measuring cylinder) of the concentrated
acid into a 1 dm^ graduated flask and making up with de-ionised water.

Nitric acid, concentrated, A.R. or A.A.S. reagent.

Ammonium citrate. Dissolve 50 g A.R. tri-ammonium citrate in 50 cm^ of
concentrated ammonia solution added with care. Cool, and make up to lOOcm^
with de-ionised water.

Ascorbic acid. Dissolve 20 g of the A.R. solid in 100 cm^ of de-ionised water.
This reagent must be freshly preparedl

Potassium cyanide. (CAUTION!) Dissolve 25 g of A.R. salt in 35 cm^ of de-
ionised water to which has been added 5 cm^ of concentrated ammonia solution.
Make up to 50 cm^ with de-ionised water and filter if necessary.

Sodium diethyldithiocarbamate (NaDDC). Dissolve 1 g of the A.R. solid in
50 cm^ of de-ionised water and filter if necessary. This reagent must be freshly
prepared.

840

FLAME SPECTROMETRY XXH, 24

Lead caprate. Prepare a standard stock solution by dissolving 0. 1 323 g of the
solid in 2 cm^ of naphthenic acid with warming. Add 20 cm^ of 4-methylpentan-
2-one (methyl isobutyl ketone), cool and then make up to the mark in a 100 cm
graduated flask with more of the ketone. . , r, 3 p

Procedure. Weigh accurately I g of the alloy and dissolve m 10 cm or
concentrated hydrochloric acid ; warm gently, and if necessary add concentrated
nitric acid dropwise (about 3 cm^) to assist the dissolution. When the vigorous
reaction is complete, digest the solution with gentle heat for about 15 minutes.
Cool, and if necessary filter through a Whatman No. 541 filter paper, washing the
beaker and filter paper with small portions of concentrated hydrochloric acid so
that a final volume of about 20 cm^ is attained. Transfer the solution to a 250 cm^
separatory funnel using a further 10 cm^ of concentrated hydrochloric acid to
effect a quantitative transfer. Add 50, cm^ of butyl acetate, shake for . one minute
and allow to separate ; iron and molybdenum are extracted into the organic layer.
Separate the two layers, collecting the acid layer and transferring, with the aid of
a further 1 0 cm^ of concentrated hydrochloric acid, to a clean 250 cm^ separating
funhel; extract with a 25 cm^ portion of butyl acetate. Again separate the two
layers, collecting the acid layer in a 250 cm^ beaker. •

Add cautiously (FUME CUPBOARD), and with constant stirring, lOcm^ of
the ammonium citrate solution; this will prevent the precipitation of metals
when, at a later stage, the pH value of the solution is increased. Then add 10 cm^
of the 20 per cent ascorbic acid, and adjust to pH 4 (BDH narrow range indicator
paper), by the cautious addition of concentrated ammonia solution down the
side of the beaker whilst stirring continuously. Then add 10 cm^ of the 50 per cent
potassium cyanide solution (CAUTION!) and immediately adjust to a pH of
9-10 (BDH indicator paper), by the addition of concentrated ammonia solution.

Transfer the solution to a 250 cm^ separatory funpel, rinsing out the beaker
with a little water. Add 5 cm^ of the 2 per cent NaDDC reagent and allow to
stand for one minute, and then add a lOcm^ portion of 4-methylpentan-2-
one (methyl isobutyl ketone), shake for one minute and then separate and collect
the organic layer. Return the aqueous phase to the funnel, extract with a further
10 cm^ portion of methyl isobutyl ketone, separate and combine the organic layer
with that already collected. Finally rinse the funnel with a little fresh ketone and
add this rinse liquid to the organic extract. In these operations the lead is
converted into a chelate which is extracted into the organic solvent.

In order to concentrate the lead extract, remove the lead from the organic
solvent by shaking this with three successive lOcm^ portions of the, dilute
hydrochloric acid solution, collecting the aqueous extracts in a 250 cm^ beaker.
To the combined extracts add 5 cm^ of 20 per cent ascorbic acid solution and
adjust to pH 4 by the addition of concentrated ammonia solution. Place the
beaker in a fume cupboard, add 3 cra^ of the 50 per cent potassium cyanide
solution and immediately adjust the pH to 9-10 with concentrated ammonia
solution. Transfer the solution to a 250 cm^ separatory funnel with the aid of a
little de-ionised water, add 5 cm^ of the 2 per cent NaDDC reagent, allow to stand
for one minute and then add 10 cm^ of methyl isobutyl ketone. Shake for one
minute and then separate and collect the organic phase, filtering it through a
fluted filter paper. This solution now contains the lead and is ready for the
absorption measurement.

Set up a double beam atomic absorption spectrophotometer with a lead
hollow cathode lamp and isolate the resonance line at 283.3 nm; adjust the gas

841

xxn, 25 QUANTITATIVE INORGANIC ANALYSIS

controls to give a fuel-lean acetylene— air flame in accordance with the operating
manual supplied with the instrument.

Prepare a blank solution by carrying through all the sequences of the
separation procedures using a hydrochloric acid solution to which no alloy has
been added, and then measure the absorption given by this blank solution, by a
series of standard solutions containing from 1— 10/<gPbcm ^ prepared by
suitable dilution of the lead caprate stock solution (Note 1), and finally of the
extract prepared from the sample of alloy. Plot the calibration curve and
determine the lead content of the alloy.

Note 1. If lead caprate is not available, standard lead solutions cun be
prepared from aqueous solutions containing known weights of A.R. lead nitrate
and following through the extraction procedure as detailed for the final
extraction of lead into methyl isobutyl ketone for the alloy. It should also be
noted that steps should be taken to avoid excessive inhalation of the vapour of
the methyl isobutyl ketone which can cause a headache.

xxn, 25. DETERMINATION OF CHROMIUM IN A NICKEL ALLOY
(AAS). The following details are reproduced by courtesy of Varian Associates
Ltd ; they refer to the analysis of an alloy containing approximately 1 5 per cent of
chromium.

Sample preparation. Dissolve 1 .000 g of the alloy in the form of fine turnings in
a mixture of hydrochloric acid ( 1 0 cm^) and nitric acid ( 1 0 cm^), then heat gently
until no more f^umes of nitrogen dioxide are obtained. Transfer to a PTFE beaker
and add dropwise 5cm^ of hydrofluoric acid (CAUTION!); during this
operation keep the solution cool and ensure that the temperature does not rise
above 30 °C. T ransfer the solution to a 1 00 cm^ plastic graduated flask and make
up to volume with de-ipnised water. With an alloy containing 15 per cent of
chromium the chromium concentration of the solution will be approximately
1500/igcm"^

Preparation of standard solutions. Prepare a stock chromium solution by
dissolving 1.000 g of pure chromium powder in lOcm^ ofhydrochloric acid; and
then follow the procedure used with the alloy up to the stage where the solution
has been diluted in the 100 cm^ graduated flask. The working standards must
contain the same reagents and major matrix elements as does the sample
solution, and the concentration of each component should be approximately the
same in the standard and test solutions. Thus the standard solutions must all
contain nickel (7000 pg cm ~ ^), iron ( 1 400 pg cm ' ^), cobalt (1 00 ;/g cm “ ^), silicon
(50/tgcm“^), hydrofluoric acid (5cm^ in each lOOcm^ of solution) and
hydrochloric acid (5cm^ in each lOOcm^ of solution). Appropriate dilution of
suitable aliquots of the stock chromium solution with a background solution of
the above composition can be used to give standards containing 0, 1000, 1500,
and 2000pgcm“^ of chromium.

Procedure. Set up a Varian Techtron Model AA-6 single beam atomic
absorption spectrophotometer (or an equivalent instrument) in accordance with
the details in the handbook. Set a chromium hollow cathode lamp in the
operating position, and select the resonance line at 520.8 nm (note that this is not
the main resonance line). Connect up the gas supply in accordance with the
handbook to give an acetylene-nitrous oxide flame. Then measure the
absorbance of the standard solutions and of the test solution and calculate the
chromium content of the alloy.

842

FLAME SPECTROMETRY XXn,26

XXII,26. DETERMINATION OF SULPHATE ION BY ATOMIC
ABSORPTION INHIBITION TITREMETRY. Introduction. The procedure
described below serves to demonstrate how atomic absorption spectrophoto-
metry can be adapted to the determination of selected anions such as sulphate,
phosphate and silicate by an indirect method. It must be stressed that the method
is non-specific for anion determinations, and further, all interfering ions, both
anions and cations, must be removed by a preliminary ion exchange treatment.
In pure solutions the anions referred to can be determined at very low
concentration levels. .■ -

The technique, termed Atomic Absorption Inhibition Titrimetry, was described
by Huber and his co-workers (Ref. 12). In this method, the experimental
conditions are deliberately chosen to encourage the occurrence of inhibition and
hence a relatively cool argon-hydrogen flame is used. The sample solution
containing the anion (S 04 ^ ~) to be determined is titrated with a standard cation
solution (Mg^'*'), with simultaneous aspiration of the mixed solution into the
flame. Initially the sulphate ions will cause formation of magnesium sulphate,
which at the flame temperature used does not dissociate. Hence the absorption by
Mg^'^ ions in the flame will be inhibited, and only when an excess of Mg^'*' ions
are present in the flame will its absorbance, which is continuously monitored on a
chart recorder, become a linear function of the volmne of magnesium solution
added to the test solution. The delay in attaining this linear behaviour is a
measure of the sulphate ion concentration in the test solution. .

Preparation of solutions. Prepare a magnesium stock solution by dissolving
0.829 g of dry A.R. magnesium oxide in the minimum quantity (approximately
42 cm^) of IM hydrochloric acid, then transfer to a 500 cm^ graduated flask and
make up to the mark with de-ionised water: in preparing this solution it is
important to avoid an excess of chloride ions. The working solution is prepared
by dilution of the stock solution to give a Mg^ ^ concentration of 200 pg cm "

A stock sulphate solution (100 /rgcm"^) is conveniently prepared by dilution
of a standard sulphuric acid solution.

Apparatus. The reaction vessel consists of a lOOcm^ tall form beaker fitted
with a plastic cover carrying two identical capillary tubes. One capillary is
attached to a reservoir burette containing the working solution of magnesium
ions, and the other is attached to the nebuliser tube of the atomic absorption
spectrophotometer. The beaker is mounted upon a magnetic stirrer so that the
solution can be stirred continuously. The uptake rate of the nebuliser must be
matched with the flow rate from the burette so that an essentially constant
volume is maintained in the reaction vessel; a typical flow rate is 2.4cm^ min"^.
It is therefore advantageous to use a spectrophotometer fitted with a variable
uptake nebulizer such as the Perkin Elmer Models 306 and 460. - '

The burner is adjusted to take an argon-hydrogen mixture, and a magnesium
hollow cathode lamp is placed in position and arranged to give the resonance line
of wavelength 285.2 nm. The output from the spectrophotometer is fed to a chart
recorder, which typically may be operated at 10 mV full scale deflection and at a
chart speed of 30 mm min ~ ^ .

Procedure. Pipette 50.0 cm^ of a standard sulphuric acid solution having a
sulphate ion concentration of 5 /ig cm " ^ into the reaction vessel and set the stirrer
m motion. Attach one capillary tube to the nebuliser inlet and the other to the
burette; set the nebuliser in operation and simultaneously open the tap on
the burette and start the recorder chart drive. Continue the addition of the

843

xxn, 27 QUANTITATIVE INORGANIC ANALYSIS

magnesium solution until the linear part of the absorption plot is well
established: see Fig. XXII, 16. Record the time taken from the start of the
magnesium ion addition to a pre-selected absorption value, tf in Fig. XXII, 16.

Repeat ^ the procedure using solutions with successive sulphate ion
concentrations of 10, 15, and 20 fig cm“^ and also a blank solution which is
simply 50 cm^ of de-ionised water. In each case measure the time taken from the
start of the magnesium ion addition to the attainment of the identical pre-
selected absorption value, i.e., (fy — f,-) where tj- is the time of attainment of the pre-
selected absorption, and h is the time of the start of the magnesium ion addition.
Plot a calibration graph of sulphate ion concentration (x-axis) against time
{tf — t;) in seconds.

Now pipette 50.0 cm^ of the test solution into the cell and repeat the procedure
described above to give the time (£/ — t,) required to achieve the pre-selected
absorption value. Then use the calibration curve to determine the unknown
sulphate concentration.

Fig. xxn, 16 tj = initial start point of titration

/y = final pre-selected absorbance value
Curve A = blank

Curves B, C, D, E, refer to sulphate solutions containing respectively 5, 10,
15, and 20 /;g cm ~ ^ of sulphate ion

xxn, 27. References

1. A. Walsh (1955). Spectrochiniica Acta., 7, 108.

2. C. S. Rann and A. N. Hambly (1965). Anal. Chem., 37, 879.

3. D. R. Thomerson and K, C. Thompson (1975) Chemistry in Britain. 11, 316.

4. T. S. West and X. K. Williams (1969). Analytica Chiin. Acta., 45, 27.

5. R. M. Dagnall, K. C. Thompson and T. S. West (1967). Talanta, 14, 551.

6. C. W. Frank, W. G. Schrenk and C. E. McLoan (1966). Anal. Client., 38, 1005.

7. V. A. Fassel, J. A. Rasmuson and T. G. Cowley (1968). Spectrochim. Acta., 23B, 579.

8. J. E. Allen (1969). Spectrochim. Acta., 24B, 13.

9. D. C. Manning and F. Femandex (1968). Atom. Absorption Newsletter, 1, 24.

10. D. G. Mitchell and A. Johansson (1970). Spectrochim. Acta., 25B, 175.

11. Safety Practices for Atomic Absorption Spectrophotometers. International
Laboratory, 1974, May/June, 63. International Scientific Communications Inc,
Fairfield, Conn.

12. C. O. Huber and R. W. Looyinga (1971). Ana/. C/ieni., 43, 498.

13. R. M. Dagnall, K. C. Thompson, and T. S. West (1967). The Analyst, 92, 506.

14. W. R. Hatch and W. L. Ott (1968). Anal. Client., 40, 2085.

844

FLAME SPECTROMETRY XXII, 28

XXn, 28. Selected bibliography

1. W. T. Elwell and J. A. F. Gidley (1966). Atomic Absorption Spectrophotometry. 2nd
edn. Oxford; Pergamon Press.

2. J. W. Robinson (1975). Atomic Absorption Spectroscopy. 2nd edn. New York; Marcel
Dekker.

3. W. Slavin (1968). Atomic Absorption Spectroscopy. New York; Interscience.

4. R. J. Reynolds, K. Aldous and K. C. Thompson (1970). Atomic Absorption
Spectroscopy. London; Griffin.

5. '^.}.VncQ{\912). Analytical Atomic Absorption Spectrometry. London; Heyden.

6. J. A. Dean (I960). Flame Photometry. New York; McGraw-Hill.

7. J. A. Dean and T. C. Rains (eds.). Flame Emission and Atomic Absorption
Spectrometry. Vol. 1 : Theory (1969); Vol. 2: Components and Techniques (1971);
Vol. 3 : Elements and Matrices (1974). New York; Marcel Dekker.

8. B. V. L’vov. (translated from the Russian by J. H. Dixon) (1970). Atomic Absorption
Spectrochemical Analysis. London; Adam Hilger Ltd.

9. J. Ramirez-Munoz (1968). Atomic Absorption Spectroscopy. Amsterdam; Elsevier.

10. G. F. Kirkbright and M. Sargent (1975). Atomic Absorption and Fluorescence
Spectroscopy. London ; Academic Press.

11. R. Mavrodineau and H. Briteux (1965). Flame Spectroscopy. New York;
Wiley-Interscience.

845

PARTG THERMAL IVIETHODS

847

CHAPTER XXlll THERMAL ANALYSIS

XXIII, 1. GENERAL DISCUSSION. Thermal methods of analysis may be
defined as those techniques in which changes in physical and/or chemical
properties of a substance are measured as a function of temperature. Methods
that involve changes in weight or changes in energy come within this definition.

Other thermal analytical techniques such as dilatometry (in which changes in
dimensions of a substance are measured as a function of temperature), or evolved
gas analysis (where qualitative and quantitative evaluations of volatile products
formed during thermal analysis are made), are outside the range of this book.

The thermal analytical techniques discussed in this chapter are;

Thermogravimetry (TG), a technique in which a change in the weight of a
substance is recorded as a function of temperature or time.

Differential Thermal Analysis (DTA), which is a method for recording the
difference in temperature between a substance and an inert reference material as a
function of temperature or time.

Differential Scanning Calorimetry (DSC), a method whereby the energy
necessary to establish a zero temperature difference between a substance and a
reference material is recorded as a function of temperature or time.

XXIII, 2. THERMOGRAVIMETRY (TG). Introduction. The basic
instrumental requirement for thermogravimetry is a precision balance with a
furnace programmed for a linear rise of temperature with time. The results may be
presented as, (i) a thermogravimetric (TG) curve, in which the weight change is
recorded as a function of temperature or time, or (ii) as a derivative
thermogravimetric (DTG) curve where the first derivative of the TG curve is
plotted with respect to either temperature or time.

A typical thermogravimetric curve, for copper sulphate pentahydrate
CuSG^.SHjO, is given in Fig. XXIII, 1.

The following features of the TG curve should be noted :

(a) the horizontal portions (plateaus) indicate the regions where there is no
weight change;

(b) the curved portions are indicative of weight losses;

wl since the TG curve is quantitative, calculations on compound stoichiometry
can be made at any given temperature.

of d^ ^XIII, 1 shows, copper sulphate pentahydrate has four distinct regions

849

XXni, 2 QUANTITATIVE INORGANIC ANALYSIS

Fig.XXra,!

CUSO4.5H2O — CUSO4.H2O
CUSO4.H2O — <• CUSO4

CUSO 4 — ► CuO -E SO 2 ■E 2 O 2

2CuO — Cu20 + i02

Approximate
temperature region
90-150 °C
200-275 °C
700-900 °C
1000-1100 °C

The precise temperature regions for each of the reactions are dependent upon the
experimental conditions (see Section XXIII, 5). Although in Fig. XXIII, 1 the
ordinate is shown as the percentage weight loss, the scale on this axis may take
other forms;

1 . as a true weight scale;

2 . as a percentage of the total weight;

3. in terms of molecular weight units.

850

THERMAL ANALYSIS XXffl,2

An additional feature of the TG curve (Fig. XXIII, 1) should now be examined,
namely the two regions B and C where there are changes in the slope of the weight
loss curve. If the rate of change of weight with time dW /At is plotted against
temperature, a derivative thermogravimetric (DTG) curve is obtained (Fig.
XXIII, 2). In the DTG curve when there is no weight loss then dWIdt = 0. The
peak on the derivative curve corresponds to a maximum slope on the TG curve.
When dWjdt is a minimum but not zero there is an inflection, i.e., a change of
slope on the TG curve. Inflections B and C on Fig. XXIII, 1 may imply the
formation of intermediate compounds. In fact the inflection at B arises from the
formation of the trihydrate CUSO4.3H2O, and that at point C is reported by
Duval (Ref. 1) to be due to formation of a golden yellow basic sulphate of
composition 2CUO.SO3. Derivative thermogravimetry is useful for many
complicated determinations and any change in the rate of weight loss may be
readily identified as a trough indicating consecutive reactions; hence weight
changes occurring at close temperatures may be ascertained.

Experimental factors. In the previous section it was stated that the precise
temperature regions for each reaction of the thermal decomposition of copper
sulphate pentahydrate is dependent upon experimental conditions. When a
variety of commercial thermobalances became available in the early 1960s it was
soon realised that a wide range of factors could influence the results obtained.
Reviews of these factors have been made by Simons and Newkirk (Ref. 2) and by
Coats and Redfern (Ref. 3) as a basis for establishing criteria necessary to obtain
meaningful and reproducible results. In addition, several sources of error can
arise in thermogravimetry which may lead to both inaccurate temperatures and
weight change values. This may necessitate the construction of a correction curve.
It must be stressed that with some modern instruments (Section XXm, 3) the
need for corrections is minimised. However, at the time of writing many
laboratories still employ thermobalances of earlier designs and in these cases a
correction curve must be constructed.

Correction curve. When an empty crucible is heated from , ambient
temperature to, say, 1000 °C there is an apparent gain in weight. This weight gain
is governed by the heating rate employed and by the crucible weight and volume.
A typical correction curve is shown in Fig. XXIII, 3. In this experiment a platinum

851

XXra, 2 QUANTITATIVE INORGANIC ANALYSIS

crucible of 1 g weight showed an apparent gain of 1.5 mg-when heated at 4 °C
min “ ‘ from ambient temperature to 1000 °C. Although the error produced is 0.1 5
per cent of the crucible weight, a 100 mg sample contained in this crucible would
suffer an apparent weight change of 1.5 per cent. This apparent weight change is
due to a variety of factors including the air buoyancy and convection currents
within the furnace. A correction curve must be constructed giving the apparent
weight change in order to calculate the actual change occurring in a sample.

The factors which may affect the results can be classified into the two main
groups of instrumental effects and the characteristics of the sample;

Instrumental factors

(a) Heating rate. When a substance is heated at a fast heating rate, the
temperature of decomposition will be higher than that obtained at a slower rate
of heating. The effect is shown for a single-step reaction in Fig. XXIII, 4. The curve

AB represents the decomposition curve at a slow heating rate, whereas the curve
CD is that due to the faster heating rate. If 7^ and Tc are the decomposition
temperatures at the start of the reaction and the final temperatures on completion
of the decomposition are Tg and 7^; the following features can be noted:

T^<Tc

TB<Tg

Tb-T^<To-Tc

The heating rate has only a small effect when a fast reversible reaction is
considered. The points of inflection B and C obtained on the thermogravimetric
curve for copper sulphate pentahydrate (Fig. XXIII, 1) may be resolved into a
plateau if a slower heating rate is used. Hence the detection of intermediate
compounds by thermogravimetry is very dependent upon the heating rate
employed.

(b) Furnace atmosphere. The nature of the surrounding atmosphere can
have a profound effect upon the temperature of a decomposition stage. For
example, the decomposition of calcium carbonate occurs at a much higher
temperature if carbon dioxide rather than nitrogen is employed as the
surrounding atmosphere. Normally the function of the atmosphere is to remove
the gaseous products evolved during thermogravimetry, in order to ensure that
the nature of the surrounding gas remains as constant as possible throughout the

852

THERMAL ANALYSIS XXm,3

experiment. This condition is achieved in many modern thermobalances by
heating the test sample m uflcuo.

The most common atmospheres employed in thermogravimetry are:

1. ‘static air’ (air from the surroundings flows through the furnace);

2. ‘dynamic air’, where compressed air from a cylinder is passed through the
furnace at a measured flow rate ;

3. nitrogen gas (oxygen free) which provides an inert environment.

Atmospheres that take part in the reaction — for example, humidified air — have
been used in the study of the decomposition of such compounds as hydrated
metal salts.

Since thermogravimetry is a dynamic technique, convection currents arising in
a furnace will cause a continuous change in the gas atmosphere. The exact nature
of this change further depends upon the furnace characteristics so that widely
differing thermogravimetric data may be obtained from different designs of
thermobalance.

(c) Crucible geometry. The geometry of the crucible can alter the slope of
the thermogravimetric curve. Generally, a flat, plate-shaped crucible is preferred
to a ‘high form’ cone shape because the diffusion of any evolved gases is easier
with the former type.

Sample characteristics

The weight, particle size, and the mode of preparation (the pre-history) of a
sample all govern the thermogravimetric results. A large sample can often create
a deviation from linearity in the temperature rise. This is particularly true when a
fast exothermic reaction is studied; for example, the evolution of carbon
monoxide during the decomposition of calcium oxalate to calcium carbonate. A
large volume of sample in a crucible can impede the diffusion of evolved gases
through the bulk of the solid large crystals especially those of certain metallic
nitrates which may undergo decrepitation (‘spitting’ or ‘spattering’) when heated.
Other samples may swell, or foam and even bubble. In practice a small sample
weight with as small a particle size as practicable is desirable for
thermogravimetry.

Diverse thermogravimetric results can be obtained from samples with different
pre-histories; for example, TG and DTG curves showed that magnesium
hydroxide prepared by precipitation methods has a different temperature of
decomposition from that for the naturally occurring material (Ref. 4). It follows
that the source and/or the method of formation of the sample should be
ascertained.

pin, 3. INSTRUMENTATION FOR THERMOGRAVIMETRY.
L^ukaszewski and Redfern (Ref. 5) outlined the following criteria for good
thermobalance design;

(a)

(b)

(c)

(d)

(e)

Tp thermobalance should be capable of continuously registering the weight
change of the sample studied as a function of temperature and time.

The furiiace should reach the maximum desired temperature (with
commercial thermobalances this can be about 1500 °C).
fhe rate of heating is linear and reproducible.

fhe sample holder should be in the hot zone of the furnace and this zone
^ould be of uniform temperature.

he thermobalance should have facilities for the provision of variable

853

XXm, 3 QUANTITATIVE INORGANIC ANALYSIS

heating rates, to permit heating in a variety of controlled atmospheres and
for heating in vacuo. The instrument should also be capable of carrying out
accurate isothermal studies.

(/) The balance mechanism should be protected from the furnace and from the
effect of corrosive gases.

(g) The temperature of the sample must be measured as accurately as possible.

(h) A balance sensitivity suitable for studying small sample weights is necessary.
An additional requirement is a facility for rapid heating and cooling of the

furnace to permit several TG analyses to be carried out in a relatively short period
of time.

Fig.XXIII,5

Apparatus. A wide range of commercial instruments is available and these
have many similar features. In this text only one instrument, the Stanton Redcroft
TG-750 Thermobalance, will be described.

The complete balance and furnace assembly of the TG 750 is shown in Fig.
XXIII, 5. The electronic microbalance B is housed in a glass bottle. The balance
has a capacity of 1 g with a switched range of sensitivities from 1 to 250 mg full
scale deflection. The sample crucible S is suspended in a platinum-rhodium
stirrup attached to the beam by an aluminium tube N.

The suspension passes through a narrow bore glass tube H with a glass flange F
at one end. The furnace assembly, C, can be raised or lowered mechanically, and
seats against F with an O-ring making a complete seal. The gas and water flow
paths are also shown in the diagram. The system may be evacuated and then
flushed with an inert gas. Access to the reference pan is obtained by removing the
glass cap G.

A diagram of the cross-section of the furnace is given in Fig. XXIII, 6. The
furnace, F, is approximately 12 mm in diameter and 20 mm tong, and the furnace
case, C, is water cooled by means of vertical channels. A platinum crucible, S, is

854

THERMAL ANALYSIS XXm,4

used to contain the sample and is heated in the furnace. Measurement of the
sample temperature is by means, of. a , platinum vs platinum-rhodium
thermocouple, T, positioned immediately below the sample crucible.

. Gas
flow ■

; X \

Fig.XXIir,6

The TG 750 operates over the range ambient temperature to 1000 °C and
heating rates from 1-100 °C min “ ^ may be employed. The furnace can be cooled
from 1000 °C to 50 °C in four minutes. Incorporation of a sensitive electronic
microbalance allows the TG 750 to measure small sample weights (1-10 mg). This
is a great advance over earlier thermobalances which used sample weights of
between 50 and 200 mg and required slow heating rates to facilitate good
resolution. This, coupled with the slow cooling rates of conventional furnaces,
meant that thermogravimetry was formerly a lengthy process.

The correction curve, described in Section XXIII, 2, due to the so-called
buoyancy effect’, has to be plotted for instruments of earlier designs. Hence the
results of TG runs require replotting before graphs of percentage weight loss
versus temperature can be obtained on older instruments.

The important features of the TG 750 are that fast heating rates may be
employed with good resolution because only small sample weights are used. The
cooling time between experiments is a matter of a few minutes. In addition,
uoyancy effects’ are reduced to a minimum so that it is possible to obtain a
irect reading of weight changes without any recourse to prior correction. The
suited for isothermal studies, and the furnace can be held at
1 balance drift. In addition to TG traces, the TG 750 will also

P 0 TG curves which are useful for the resolution of weight changes occurring
at temperatures close to each other.

of THERMO gravimetry. Some of the
These are Ibermogravimetry are of particular importance to the analyst.

t^^l^rmination of the purity and thermal stability of both primary and

secondary standards. r f r

855

XXin, 4 QUANTITATIVE INORGANIC ANALYSIS

2. The investigation of correct drying temperatures and the suitability of various

weighing forms for gravimetric analysis.

3. Direct application to analytical problems (automatic thermogravimetric

analysis).

4. The determination of the composition of complex mixtures.

Thermogravimetry is a valuable technique for the assessment of the purity of

materials. Analytical reagents, especially those used in titrimetric analysis as
primary standards, e.g., sodium carbonate, sodium tetraborate, and potassium
hydrogenphthalate, have been examined. Many primary standards absorb
appreciable amounts of water when exposed to moist atmospheres. TG data can
show the extent of this absorption and hence the most suitable drying
temperature for a given reagent may be determined.

The thermal stability of EDTA as the free acid and also as the more widely used
disodium salt, Na2EDTA.2H20, has been reported by Wendlandt (Ref. 6). He
showed that the dehydration of the disodium salt commences at between 110 and
125 °C, which served to confirm the view of Blaedel and Knight (Ref. 7) that
Na2EDTA.2H20 could be safely heated to constant weight at 80 °C.

Undoubtedly the most widespread application of thermogravimetry in
analytical chemistry has been in the study of the recommended drying
temperatures of gravimetric precipitates. Duval studied over a thousand
gravimetric precipitates by this method and gave the recommended drying
temperatures. He further concluded that only a fraction of these precipitates are
suitable weighing forms for the elements. The results recorded by Duval were
obtained with materials prepared under specified conditions of precipitation and
this must be borne in mind when assessing the value of a given precipitate as a
weighing form, since conditions of precipitation can have a profound effect on the
pyrolysis curve. It must be stressed that the rejection of a precipitate because it
does not give a stable plateau on the pyrolysis curve at one given rate is
unjustified. Further, the limits of the plateau should not be taken as indicative of
thermal stability within the complete temperature range. The weighing form is
not necessarily isothermally stable at all temperatures that lie on the horizontal
position of a thermogravimetric curve. A slow rate of heating is to be preferred,
especially with a large sample weight, over the temperature ranges in which
chemical changes take place. Thermogravimetric curves must be interpreted with
due regard to the fact that whilst they are being obtained the temperature is
changing at a uniform rate, whereas in routine gravimetric analysis the
precipitate is often brought rapidly to a specified temperature and maintained at
that temperature for a definite time.

Thermogravimetry may be used to determine the composition of binary
mixtures. If each component possesses a characteristic unique pyrolysis curve,
then a resultant curve for the mixture will afford a basis for the determination of
its composition. In such an automatic gravimetric determination the initial
weight of the sample need not be known. A simple example is given by the
automatic determination of a mixture of calcium and strontium as their
carbonates.

Both carbonates decompose to their oxides with the evolution of carbon
dioxide. The decomposition temperature for calcium carbonate is in the
temperature range 650-850 °C, whilst strontium carbonate decomposes between
950 and 1150 °C. Hence the amount of calcium and strontium present in a
mixture may be calculated from the weight losses due to the evolution of carbon

856

THERMAL ANALYSIS XXm,5

dioxide at the lower and higher temperature ranges respectively. This method
could be extended to the analysis of a three-component mixture, as barium
carbonate is reported to decompose at an even higher temperature (~ 1300 °C)
than strontium carbonate.

A further example, cited by Duval (Ref. 8), is the automatic determination of a
mixture of calcium and magnesium as their oxalates. Calcium oxalate
monohydrate has the following three distinct regions of decomposition;

temperature range °C
100-250
400-500
650-850

In comparison with this, magnesium oxalate dihydrate has only two
decomposition stages:

temperature range °C

(d) MgC204.2H20 -^MgC204 + 2H20 100-250

(e) MgC^O^ — ^MgO-l-CO-t-COi 400-500

A pyrolysis curve for a mixture of these two oxalates would thus show three
decomposition steps. The final step would be due entirely to the loss of carbon
dioxide from calcium carbonate and hence the amount of calcium present in the
mixture may be calculated. The amount of magnesium in the oxalate mixture
may be calculated from the second step (at which the stages (h) and (e) occur}
because the amount of carbon monoxide due to calcium carbonate may be
subtracted from the total observed weight loss, and the remainder is thus due to
the loss of carbon dioxide and carbon monoxide from anhydrous magnesium
oxalate.

Complex materials (for example, clays and soils) have been the subject of
thermogravimetric study by Hoffman et al. (Ref. 9). The pyrolysis curves of
most soils examined showed plateaus starting between 150-180 °C and
extending to 210-240 °C, indicating that hygroscopic moisture and/or easily
volatile organic compounds had been removed. When the clay content of a soil
was studied the loss in weight at 500 °C read from a pyrolysis curve gave an
estimate of the organic matter which was in reasonable agreement with dry
combustion and wet oxidation data. An additional feature of the work suggests
that lattice water may be quantitatively determined in pure clays. Because lattice
water came off from different clays at different temperatures, these temperatures
may possibly be used as a method of identification.

(a) CaC204.2H20 CaC204 +2H2O
(h) ,CaC204 ^ CaCOa + CO

5. EXPERIMENTAL. A limited number of thermogravimetric
experiments will be outlined below.- For more detailed information on these and
other studies the reader is referred to the publications listed in the bibliography at

the end of this chapter.

^en using a modern thermobalance, e.g., the Stanton Redcroft TG-750,
jy ™ incorporates an electronic microbalance requiring small sample weights,
t i pperating precautions should be noted:
antic' ^ j sample selected is dependent upon the actual weight loss

(b) The crucible should not be handled, because of the danger of transferring
it moisture to the crucible. A platinum crucible may be cleaned by placing

clea d platinum crucible is heavily contaminated it may be

ned by heating with a sodium carbonate/sodium nitrate fusion mixture.

857

XXffl, 5 QUANTITATIVE INORGANIC ANALYSIS

material is thought to be inhomogeneous, several samples should be run and
different results will confirm the inhomogeneity. The sample particle size should
be smaller than 100 mesh ( < 150//m) to ensure that an even layer is distributed in
the crucible. ■ . . . r •

(d) The method of obtaining a sample depends upon the nature of the material;
thus a circular disc may be cut from a film of material by the use of an appropriate
cork borer or leather punch. Fibrous material, which does not pack easily, may be
squeezed between metal foil before being transferred to the crucible. Liquid
samples may be transferred to the crucible by means of a hypodermic syringe:
Air-sensitive samples should be loaded on to the crucible in a glove box and
transferred rapidly to the thermobalance which should be set up all ready for a
dry inert gas flow. Materials which creep or froth should not be used in a
thermobalance. It is always sound practice to heat the test material in a small
crucible in an oven or muffle furnace to ascertain whether or not there is any
creeping or frothing before using the sample for thermogravimetry. Considerable
damage can occur to a thermobalance if samples are not monitored in this way
prior to analysis in the apparatus.

The following experiments are designed to make the operator familiar with the
use of the thermobalance:

A. The thermal decomposition of calcium oxalate monohydrate. This
determination may be carried out on any standard thermobalance. In all cases
the manufacturer’s handbook should be consulted for full detailed instructions
for operating the instrument.

Initially, zero the balance on the 10 mg range (in the case of the Stanton
Redcroft TG-750) with an empty crucible in position and use an air flow of 10 cm^
min“‘. Weigh accurately about 2 mg of the calcium oxalate monohydrate
directly into the crucible and record the weight on the chart. The recorder
variable range may now be used to expand the sample weight to 100 per cent of
full scale. Select a suitable heating rate (30 °C min“‘) and record the pyrolysis
curve of calcium oxalate monohydrate from ambient temperature to 1000 °C in
terms of percentage sample weight loss. From the TG curve estimate the purity of
the calcium oxalate (see Section XXIII, 4).

As additional experiments, investigate the decomposition of calcium oxalate in
a static air atmosphere and in a nitrogen atmosphere at a flow rate of 10 cm^
niin“^. Compare the final stage of the decomposition, i.e., the conversion of
calcium carbonate to calcium oxide, using different furnace atmospheres.

B. The thermal decomposition ofcopper sulphate pentahydrate. Follow the
procedure outlined in A above, but in this case weigh out accurately about 6 mg of
the copper sulphate. Record the thermal decomposition of copper sulphate from
ambient temperature to 1000 °C using a heating rate of 10°C min“* and an air
atmosphere with a flow rate of 10 cm^ min“ Examine the effect of varying the
heating rate on the dehydration reactions by selecting rates of 2, -20 and
100 °C min“' in addition to 10 °C min“‘ rate used previously. Further experi-
ments may be designed to study the effect of differing particle size on the
dehydration reactions of copper sulphate pentahydrate.

C. Other useful substances to study. The following substances show
interesting pyrolysis curves and an assessment of the purity of these materials
may be investigated:

cadmium sulphate, SCdSO^SHjO ;

858

THERMAL ANALYSIS XXin,6

ammonium magnesium phosphate, iVIgNH4P04.6H20;
and the disodium salt of ethylenediaminetetraacetic acid, Na2EDTA.2H20. >

The automatic gravimetric determination of calcium and. magnesium as their
oxalates, outlined in Section XXin, 4, is an obvious extension of experiment A
above and may be readily carried out. The reader should be aware that if an early
design of thermobalance requiring a sample weight of several hundred milligrams
is used a correction curve (Section XXIQ, 2) must be apphed before determining
the calcium and magnesium present in the mixture.

XXffl, 6. DIFFERENTIAL THERMAL ANALYSIS AND DIFFEREN-
TIALSCANNING CALORIMETRY. Introduction. In differential thermal
analysis (DTA) both the test sample and an inert reference material (usually
a alumina) undergo a controlled heating or cooling programme which is usually
linear with respect to time. There is a zero temperature difference between the
sample and the reference material when the former does not undergo any
chemical or physical change. If, however, any reaction takes place, then a
temperature difference AT will occur between the sample and the reference
material. Thus in an endothermic change, e.g., when the sample melts or is
dehydrated, the sample temperature is lower than that of the reference material.
This condition is only transitory because on completion of the reaction the
sample will again show zero temperature difference compared with the reference.

In DTA a plot is made of AT against temperature or time, if the heating or
cooling programme is linear with respect to time. An idealised DTA curve is
shown in Fig. XXIII, 7 , in which ( 1 ) is an exothermic peak and ( 2 ) is an

cn othermic peak. Both the shape and size of the peaks can give a large amount of
m ormation about the nature of the test sample. Thus sharp endothermic peaks
changes in crystallinity or fusion processes, whereas broad
^^ise from dehydration reactions. Physical changes usually result in
n othermic curves whilst chemical reactions, particularly those of an oxidative
?'^^‘^°“inantly exothermic.

. scanning calorimetry (DSC) measures the differential energy

quired to keep both the sample and reference chemicals at the same

859 •

XXffl, 7 QUANTITATIVE INORGANIC ANALYSIS

temperature. Thus, when an endothermic transition occurs, the energy absorbed
by the sample is compensated by an increased energy input to the sample in order
to maintain a zero temperature difference. Because this energy input is precisely
equivalent in magnitude to the energy absorbed in the transition direct
calorimetric measurement of the energy of the transition is obtained from this
balancing energy. The DSC curve is recorded with the chart abscissa indicating
the transition temperature and peak area measures the total energy transfer to or.
from the sample. - •

XXni, 7. INSTRUMENTATION FOR DTA AND DSC. A block diagram
of a differential thermal analyser is shown in Fig. XXIII, 8. The basic instrument
consists of;

a sample and reference holder assembly;
furnace control;

a reaction chamber allowing analysis in a variety of atmospheric systems;
a suitable sensor to measure the temperature difference between the sample
and reference material;
an amplifier for AT;
a suitable chart recorder.

FourDTAsystemsareofiferedby
Stanton Redcroft, operating over
a variety of temperat ure ranges.

The sample platforms and heat-
er assembly of the Stanton
Redcroft DTA 671B are shown in
Fig. XXIII, 9. Two small metal
platforms, (1) and (2), are each
welded to a chrome-alumel
thermocouple, (3) and (4). The
platforms are moimted in the base
of a hollow metal cup (5), the walls
of which are wound externally

Fig.XXin, 8 with a specially insulated heater

(6). The metal disc (7) is above the
base of the cylindrical cup through which the thermocouple assemblies protrude.
The sample dishes are in good thermal contact with the platforms. The gas inlet
(8) to the sample chamber (9) is via a small capillary tube (10).

The heater assembly is fitted
with a tight fitting metal lid which
may incorporate either a Pyre.x
window or be fitted with a small
capillary with one end in position
over the sample platform. The
total volume of the sample
chamber is about 5 cm^.

Sample containers used with
this instrument are dish shaped,
and are usually made of either
aluminium or platinum. The speci-
men dishes have very flat bases

860

THERMAL ANALYSIS XXin,8

to give good thermal contact between them and the therniocouple platform.

The Perkin Elmer Differential Scanning Calorimeter Model DSC-2 is shown
in Fig. XXIII, 10.

In this instrument the sample and reference holders are identical in all respects ;
both have a built-in heater and temperature sensor. The holders must be
thermally and mechanically stable and chemically inert over the entire
temperature range of the instrument ( - 175 °C to 725 °C).

The structure of the holder is in the form of a partially hollowed out cylinder in
which the sample or reference pan is placed. Below the base of the cylinder is a
platinum resistance thermometer and a platinum wire heating element. The
platinum temperature sensors detect the slightest fluctuation of the sample
holder temperature compared to the reference holder temperature caused by
evolution or absorption of energy by the sample. The electronic system provides
differential electrical power to the heaters to compensate exactly for the
fluctuation and to maintain a null balance condition. The differential power is
read out directly in millicalories per second on the recorder and is equivalent to
the rate of energy of sample absorption or evolution. The model DSC-2 operates
normally over the range 50° C to 725 °C, but with a sub-ambient accessory the
®t^closure can be cooled to work down to — 175 °C.

Eleven different heating rates varying from 0.3125 °C to 320 °C min"^ are
available from the temperature programmer. The same variety of cooling rates
are obtainable from the instrument.

8- experimental and instrumental factors, dta

y'L peaks are also governed by the factors affecting TG curves (Section
I Hence, heating rates, atmosphere, and geometry of sample holders can

alter the position of DTA and DSC peaks.

tnost important factor in obtaining reliable results for both
sho preparation of the sample and reference material. Great care

^ e taken in the preparation of the sample and in the way the crucible or

861

XXm, 9/10 QUANTITATIVE INORGANIC ANALYSIS

ampoule is loaded. To obtain, reproducible results between successive
experiments it is essential to ensure that precisely the same packing procedure is
carried out each time. The selection and handling of samples for DTA is similar to
that outlined in Section XXm, 5 (c) and (d). It is possible, however, to use
materials which creep, froth or boil if sealed sample containers are used to ensure
no damage occurs to the sample holder assembly. With most modern DTA
apparatus a device for encapsulation of the sample is available. In the Perkin
Elmer Model DSC-2 it is usual practice to encapsulate the sample in metal pans
of high thermal conductivity, to ensure that the sample is in the form of a thin
wafer which enables the best thermal contact between sample and temperature
sensor.

It is now standard practice to use an empty pan as the reference in DSC (a
similar practice is made in DTA when the sample weight is of the order of 1 mg).
With higher sample weights it is necessary to use a reference material, for, ideally,
the total weight of the sample and its container should be approximately the same
as that of the reference and its container. The reference material should be
selected so that it possesses similar thermal characteristics to the sample. The most
widely used reference material is a alumina which must be of analytical reagent
quality. Before use, a alumina should be recalcined and stored over magnesium
perchlorate in a desiccator. Kieselguhr is another reference material normally
used when the sample is of a fibrous nature. If there is an appreciable difference
between the thermal characteristics of the sample and reference materials, or if
values of AT are large, then dilution of the sample with the reference substance is
sensible practice. Dilution may be accomplished by thoroughly mixing suitable
proportions of sample and reference material.

XXni, 9. APPLICATIONS OF DIFFERENTIAL THERMAL ANALYSIS
AND DIFFERENTIAL SCANNING CALORIMETRY. DTA and DSC
may both be used in conjunction with TG for certain analytical applications, e.g.,
the determination of moisture content or the analysis of solid mixtures.

Early applications of DTA were in the qualitative analysis of complex
materials. Thus DTA provided a rapid method for the ‘fingerprinting’ of
minerals, clays, and polymeric materials. Indeed, an extremely wide range of
materials may be studied by DTA and DSC. The areas of study include thermal
stability and decompositions, fusion, phase changes, and purity determinations.
A recent important application has been in the measurement of the degree of
conversion of high alumina cement, details of which are outlined below.

It must be stressed that all the thermal methods outlined in this chapter are
frequently used in conjunction with other techniques. Thus the analysis of
evolved gases during a TG, DTA or DSC experiment may be performed by gas
chromatography or mass spectrometry. X-ray crystallography may be used to
study the structure of reaction intermediates isolated as a result of thermal
studies.

XXIII, 10. EXPERIMENTAL. In this section only one experiment using
DTA and one using DSC will be outlined. For detailed information on the
considerable number of analyses performed by both of these techniques the
reader is referred to the bibliography at the end of this chapter.

(A) DTA studies of copper sulphate pentahydrate. In order to become
familiar with the use of the instrument this experiment may be carried out on any

862

THERMAL ANALYSIS XXffl, 10

standard differential thermal analyser. The manufacturer’s handbook should be
consulted for detailed instructions on instrument operation.

The dehydration and decomposition, peaks of CUSO4.5H2O may be
compared with those obtained by the TG determination (Section XXIH, 5 , B).

In the case of the Stanton Redcroft DTA 673 dr 674 the following experimental

conditions may be used:

Procedure. Weigh accurately the pair of empty crucibles and record
each individual weight. Prepare the sample by mixing together equal weights of
powdered copper sulphate pentahydrate and a alumina (obtainable from BDH
Ltd, Poole, Dorset). Weigh out accurately about 60 mg of the diluted sample into
one of the previously- weighed crucibles. Load the other crucible with a alumina
so that the combined weight of this reference crucible plus a alumina is equal to
the weight of the sample crucible and the diluted sample. Insert the crucibles,
using tweezers, carefully into the thermocouple wells ensuring that they are
correctly located (the reference crucible goes into the front well). Lower the
furnace over the sample holder assembly and locate in the furnace mounting.
Select the appropriate sample atmosphere, e.g., flowing dry air, at a heating rate
of 10 °C min“^ Choose the appropriate amplifier range, chart speed and
temperature programme, from ambient temperature to 1000 °C (for full details
see the manufacturer’s handbook). Record the DTA plot for copper sulphate
pentahydrate over the desired temperature range.

To gain further experience with the use of the instrument the dehydration
studies of CUSO4.5H2O (up to about 550 ° C) may be performed using, (i)
different heating rates, (ii) various gas atmospheres, (iii) samples of different
particle size.

Materials that undergo crystalline transition and fusion may be useful for
alternative DTA studies. Suitable substances include potassium nitrate, sodium
chromate, and potassium sulphate.

(B) The determination of the degree of conversion of high alumina
cement Introduction. Both DTA and DSC instruments have been used for the
study of conversion of high alumina cements (Ref. 14 ). In the experiment
described below the Perkin Elmer Model DSC -2 Differential Scanning
Calorimeter was employed.

Theory. High alumina cements undergo a ‘conversion’ reaction whereby
the metastable compounds CaO. AI2O3. IOH2O, 2CaO.AI2O3.8H2O and
alumina gel of the early set cement are converted to more stable materials
represented by the following reactions;

(a) 3 (CaO . AI2O3 . IOH2O) 3 CaO . AI2O3 . 6H2O + 2 (Al 203 . 3H2O)

3 ( 2 CaO . AI2O3 . 8H2O) 2 ( 3 CaO . AI2O3 . 6H2O) + AI2O3 . 3H2O
AI2O3 . XH2O — ^ AI2O3 . 3H2O

Only small quantities of 2 CaO . AI2O3 . 8H2O are usually formed during the
the cement, and the alumina gel initially formed disappears after a
in determine the degree of conversion the quantities of the materials

be H ^fluations must be estimated. The degree of conversion may

3/-, as follows: half conversion is when the quantity of

staee f ' ®2C) is equal to the quantity of CaO . AI2O3 . IOH2O. At the
° approximately 50 per cent conversion the DSC peaks are found to be of

863

XXra, 11 QUANTITATIVE INORGANIC ANALYSIS

equal size. A third peak due to AljOj . 3H2O is also present, and it is found to be
of an equivalent size.

The degree of conversion may be obtained from the relationship :
amount of 3CaO . AI2O3 . 6H2O x 100
amount of 3CaO . AI2O3 . 6H2O + amount of CaO . AI2O3 . IOH2O

When cement is exposed naturally a carbo-aluminate 3CaO.Al203.
CaC03 . aq. may be formed due to the carbonation of 3CaO . AI2O3 . dHjO and
this could result in a small apparent degree of conversion. Because of this fact and
since AI2O3 .3H2O is not decomposed the progress of the reactions shown by
equations (a) and (c) may be determined by the following relationship;

amount of AI2O3 . 3H2O x 100

amount of AI2O3 . 3H2O + amount of alumina gel 4- CaO . AI2O3 . IOH2O

As mentioned above, the alumina gel disappears after a few months and the
relationship becomes:

amount of AI2O3 . 3H2O x 100
amount of AI2O3.3H2O + amount of CaO. AI2O3 . IOH2O

Strong endothermic peaks are obtained for CaO.Al^Oj. IOH2O at 110—120 °C,
AI2O3.3H2O at 295-310 °C, and for 3CaO.Al2O3.6H2O at 320-350 °C.
The exact value of the transition temperatures depends upon factors such as
sample size, sample packing, and heating rate. Using DTA, the percentage of the
compounds is proportional to peak height, the relationship does not hold for
DSC, and therefore it is necessary to calibrate using standard samples of high
alumina cement in which the degree of conversion is known. In this way the
following relationship is obtained:

B X peak height due to AI2O3 . 3H2O x 100
(B X peak height due to AI2O3 . 3H20)+(A x peak height due to
CaO.Al2O3.10H2O)

where ‘A’ and ‘B’ are constants obtained from the calibration standards. The
peaks chosen are the broad endotherm at 100-120 °C (CaO . AljOj . IOH2O) and
the first of the two peaks (if two peaks are present) at 300-350 °C (AI2O3 . 3H2O).

Problems may arise if other materials are present in the sample: e.g., Ettringite,
3Ca0.Al203.3CaS04..32H20, may be formed due to sulphate attack and
produces a sharp peak superimposed on the broad endotherm due to
CaO.Al2O3.10H2O.

THERIWOMETRIC TITRATIONS

XXin, 11. INTRODUCTION. Although thermometric titrimetry is not
strictly a thermal method of analysis as previously defined (Section XXIII, 1 ), the
technique merits inclusion in this part of the book. Thermometric titrations
involve the mesurement of the change in temperature of a system as a function of
time, or of the volume of titrant. Thus the temperature change of a solution is
recorded as titrant is added. The titrations are carried out under as near adiabatic

864

THERMAL ANALYSIS XXffl, 12

conditions as possible in order to minimise heat losses between the titrated
solution and its surroundings. In practice the titrant is added from an automatic
burette delivering a constant volume into a thermally insulated vessel containing
thetitrand (the solution to be titrated). A plot of temperature against volume of
titrant (or time) is made. A' thermogram of a simple acid-base thermometric
titration is shown in Fig. XXIII, 1 1 . AB corresponds to the situation before titrant

IS added and the titration vessel and its contents are gaining or losing heat to the
surrouiidings. At B the titration is started and heat is evolved, mainly from heat of
neutralisation. At C the end-point is reached and the heat changes recorded after
tnis, along CD, arise only from heats of dilution and differences in temperature
nat may exist between the titrant and titrand.

XXin, 12. THEORY. An essential condition for a successful titration is that
ere should be a change in free energy, AG® in the end-point region. This
eon ition is based on the equilibrium constants of the reactions involved.

Thus-AG® = RTlnK

A thermometric titration, however, depends upon the heat of the reaction in

accordance with the equation.

AB = AG-bTAS

Hence a thermometric titration may be feasible when there is a significant change
yield^ (providing that free energy changes are favourable), to

of a ^PP^5>able overall change in enthalpy AH. The potentiometric titration

end I - sodium hydroxide affords a scarcely discernible

whereas the thermometric titration of boric acid under the same
The'cl?^ ^ clearly marked end point (Ref. 10).

ongein temperature AT of an acid-base titration is dependent upon the

865

xxni, 13/14 QUANTITATIVE INORGANIC ANALYSIS

molar heat of neutralisation AH^ and is given by the relation:

where N is the number of moles of water formed by neutralisation and Q is the

heat capacity. Since AH„, and Q are constant throughout the titration, AT is
proportional to /V.

XXin, 13. INSTRUMENTATION. The basic instrumentation introduced
by Linde, Rogers and Hume (Ref. 11) extended the scope of thermometric
titrimetry. The titrant was added continuously by means of a motor-driven
syringe burette into a Dewar flask fitted with a mechanical stirrer. A thermistor,
which formed part of a bridge circuit, was placed in the titration vessel and acted
as the temperature sensor. Continuous recording of the unbalance potential from
the bridge circuit enabled rapid automatic titrations to be made for the first time.

A commercial instrument, the Aminco Titra-Thermo-Mat (American
Instrument Co, Silver Spring, Maryland, USA) has been developed, in which the
titration vessel is located in an insulated enclosure— the ‘adiabatic titration
tower’. The output from two thermistors, one measuring the titrant temperature
and the other the temperature of the titration vessel, is fed into a bridge circuit.

A simple apparatus, employing a single thermistor, which is suitable for
elementary studies js described by Williams (Ref. 12), in which a magnetic stirrer
mixes the solution contained in an inner beaker which is thermally insulated from
the surroundings by an outer beaker. A Perspex lid covers the beakers and
supports the thermistor and the capillary tip of a burette. A constant flow is
maintained by using a simple constant pressure head device, a separatory funnel
fitted with a capillary and screw-clip. The thermistor, Stantel F23 (supplied by
Electronic Services, Harlow, Essex) is connected into one arm of the Wheatstone
bridge circuit, shown in Fig. XXIII, 12, in which R is a resistance box.

1-5 V

Fig. xxm, 12

The titration is followed by plot-
ting the thermistor resistance against
time as follows. A given resistance
value is set on the resistance box and
the time noted when the galvan-
ometer G indicates null deflection.
The resistance is then changed to
another fixed value and the time is
noted again for null deflection. A
series of readings is obtained and a
resistance against time plot is con-
structed, from which the point of
inflection is obtained by drawing
lines through the experimental points.

XXin, 14. APPLICATIONS. Thermometric titrations may be used in
aqueous solutions to follow neutralisation, precipitation, complexation; and
redox reactions.

Neutralisation titrations. Thermometric titrations of strong acids and bases
are widely reported and may be easily performed using the instrumental methods
outlined above. However, a more significant application is the titration of weak
acids with strong bases, e.g., boric acid and sodium hydroxide. In general, O.OIM

8«6

THERMAL ANALYSIS XXIH.IS

solutions of all acids with-pX^ < 10 may be titrated thermometrically with a
precision of 1 per cent provided the heat neutraUsation is not less than
42 kJ mole"^ Mixtures of a strong acid and a weak acid can be titrated and
the resulting thermogram has two inflection points.

Precipitation titrations. When a slightly soluble compound MX is formed
heat is either evolved or' absorbed according to the equation;

MX,, ± AX

where AH is the heat of reaction. If AH is sufficiently large then the heat
change may be followed by thermometric titrimetry. Examples include
determination of halides with silver and mercury(II) ions, and the estimation of
calcium, strontium, or barium with oxalate ion.

Redox titrations. Although relatively little work has been done on redox
reactions, the technique appears to be successful for these purposes.
Thermometric titrations of permanganate with iron(II) and with oxalic acid
compare favourably with the results obtained by conventional titrimetry.

Compkxation titrations. Jordan and Alleman (Ref. 1 3) have studied the
thermometric titrations of divalent metal cations with EDTA. An interesting
application is to the titration of a binary mixture of calcium and magnesium ions
with the tetrasodium salt of EDTA. In the normal titrimetric procedure using
Solochrome Black as indicator (Section X, 55) only the total calcium and

magnesium can be evaluated because
the log stability constants of calcium
EDTA and magnesium EDTA are
10.7 and 8.7 respectively. Using a
thermometric titration the curves ob-
tained (Fig. XXIII, 13) show an exo-
thermic section AB (calcium) and an
endotherm BC (magnesium)i The endo-
thermic character of the chelation of the
magnesium ion and the exothermic
character of the chelation of calcium are
mainly attributed to a significant differ-
ence in the entropies of chelation.

®fin, 15. EXPERIMENTAL. The following experiment is illustrative of the
procedure which has to be followed for thermometric titrations and can be
irectly adapted for use in other determinations such as those mentioned in
Section XXm, 14.

The therraometric titration of chloride

described by Williams (Ref. 12) and previously outlined in
Camh^'H suitable for this titration. A Pye Scalamp galvanometer and a

fnstrument Decade Resistance Box are used in the bridge circuit
head d^^ electronic components indicated in Fig. XXIII, 12. The constant

evice IS a 250-cm^ separatory funnel fitted with a capillary and a screw clip.
rocedure. Prepare the following solutions :*
of thp 0-lM-sodium chloride solution by weighing out about 3.0 g

salt and dissolving it in 500 cm^ of water.

867

xxni, 16/17 QUANTITATIVE INORGANIC ANALYSIS

A 0.05M-silver nitrate solution by weighing out accurately 4.248 g of the salt
and dissolving it in 500 cm^ of water in a graduated flask.

Pipette 50 cm^ of the 0.05M-silver nitrate solution into a 250-cm^ beaker which
serves as the inner titration vessel and add 10 cm^ of 2.5M-nitrlc acid. Place the
250-cm^ beaker in an outer 600-cm^ beaker and enclose the whole in a block of
expanded polystyrene. Locate the thermistor into the titration cell and connect it
to the bridge circuit. Fill the constant head separatory funnel with the sodium
chloride solution and adjust the flow rate with the screw-clip to about 2.5 cm^
min~L Connect the ‘burette’ capillary tip to the titration vessel and start the
titration. Using the procedure outlined in Section XXin, 12, plot a graph of
thermistor resistance against time and determine the point of inflexion. Hence
calculate the chloride ion concentration.

XXm, 16. References

1. C. Duval and M. de Clercq (1951). Anal. Chim. Acta., 5, 282.

2. E. K. Simons and A. E. Newkirk (1964). Talanta, 11, 549.

3. A. W. Coats and J. P. Redfern (1963). Analyst, 88, 906.

4. R. C. Turner, I. Hoffman and D. Chen (1963). Can. J. Cltem., 41, 243.

5. G. M. Lukaszewski and J. P. Redfern (1961). Lab. Practice, 10, 552.

6. W. W. Wendlandt (1960). Anal. Chem., 32, 848.

7. W. J. Blaedel and H. T. Knight (1954). Anal. Chem., 26, 741.

8. C. Duval (1963). Inorganic Thermogravimetric Analysis. 2nd edn. p. 93. Amsterdam,
London and New York; Elsevier.

9. I. Hoffman, M. Schnitzer and J. R. Wright (1959). Anal. Chem., 31,440.

10. J. Jordan (1963). J. Chem. Ecluc., 40, A5.

1 1. H. W. Linde, L. B. Rogers and D. N. Hume (1953). Anal. Chem., 25, 404.

12. D. R. Williams (197 1). Education in Chemistry, 8, 97.

13. J. Jordan and T. G. Alleraan (1957). Anal. Chem., 29, 9.

14. H. G. Midgley and A. Midgley (1975). Magazine of Concrete Research, 27, 55.

XXin, 17. Selected bibliography

1. C. Duval (1963). Inorganic Thermogravimetric Analysis. 2nd edn. Amsterdam,
London and New York; Elsevier,

2. P. D. Garn (1966). Thermoanalytical Methods of Investigation, New York; Academic
Press.

3. C. J. Keatch (1975). An Introduction to Thermogravimetry. 2nd edn. London;
Heyden.

4. R. C. Mackenzie (ed.). Differential Thermal Analysis. Vol. 1 (1970), and Vol. 2 (1972).
London; Academic Press.

5. W. W. Wendlandt (1974). Thermal Methods of Analysis. 2nd edn. New York; Wiley-
Interscience.

6. R. F. Schwenker and P. D. Garn (cds.). Thermal Analysis. Vol. 1 (1969), and Vol. 2
(1968). New York; Academic Press.

7. H. J. V. Tyrrell and A. E. Beezer (1968). Thermometric Titrimetry. London;
Chapman and Hall.

8. L. G. Bark and S. G. Bark (1968). Thermometric Titrimetry. Oxford; Pergamon.

* It is sound practice to make up the sodium chloride and silver nitrate solutions and leave them in a
thermostat for 2-3 hours before proceeding with the thermometric titration. This is to ensure there
is no appreciable temperature difference between the titrant and titraiid solutions.

868

APPENDIX I INTERNATIONAL ATOMIC WEIGHTS. 1973

Element

Symbol

Atomic

No.

Atomic

Weight

Element

Symbol

Atomic

No.

Atomic

Weight

Actinium

(227)

Mercury

200.59

Aluminium

26,9815

Molybdenum

95.94

Americium

(243)

Neodymium

144.24

Antimony

121.75

Neon

20.179

Argon

39.948

Neptunium

237.0482

Arsenic

74.9216

Nickel

58.70

Astatine

(210)

Niobium

92.9064

Barium

137.34

Nitrogen

14.0067

Berkelium

(247)

Nobelium

102

(255)

Beryllium

9.0122

Osmium

190.2

Bismuth

208.9804

Oxygen

15.9994

Boron

10.81

Palladium

106.4

Bromine

79.904

Phosphorus

30.9738

Cadmium

112.40

Platinum

195.09

Calcium

40.08

Plutonium

(244)

Californium

(251)

Polonium

(209)

Carbon

12.011

Potassium

19 (

39.098

Cerium

140.12

Praseodymium

140.9077

Cesium

132.9054

Promethium

(145)

Chlorine

35.453

Protactinium

231.0359

Chromium

51.996

Radium

226.0254

Cobalt

58.9332

Radon

(222)

Copper

63.546

Rhenium

186.207

Curium

(247)

Rhodium

102.9055

Dysprosium

162.50

Rubidium

85.4678

Einsteinium

(254)

Ruthenium

101.07

Erbium

167.26

Samarium

150.4

Europium

151.96

Scandium

44.9559

Fermium

Fra

100

(257)

Selenium

78.96

Fluorine

18.9984

Silicon

28.086

Francium

(223)

Silver

107.868

Gadolinium

157.25

Sodium

22.9898

Gallium

69.72

Strontium

87.62

Germanium

72.59

Sulphur

32.06

Gold

196,9665

Tantalum

180.9479

Hafnium

178.49

Technetium

(97)

Helium

4.0026

Tellurium

127.60

Holmium

164.9304

Terbium

158.9254

Hydrogen

1.0079

Thallium

204.37

Indium

114.82

Thorium

232.0381

Iodine

126.9045

Thulium-

168.9342

Iridium

192.22

Tin

118.69

Iron

55.847

Titanium

47.90

Krypton

83.80

Tungsten

183.85

Lanthanum

138,9055

Uranium

238.029

Lawrencium

103

(260)

Vanadium

50.9414

Lead

207.2

Xenon

131.30

Lithium

6.941

Ytterbium

173.04

Lutetium

174.97

Yttrium

88.9059

Magnesium

24.305

Zinc

65.38

Manganese

54.9380

Zirconium

91.22

Mendelevium

101

(258)

Notes:

1. This table is scaled to the relative atomic mass = 12.

2. Values in parentheses refer to the isotope oflongest known half-life for radioactive elements.

3. Information provided here is based upon the Report of the Commission on Atomic Weights.
Pure and Applied Chemistry, (1974), 37, 589.

870

appendix li INDEX OF ORGANIC CHEMICAL REAGENTS

The following table is included by kind permission of Hopkin and Williams Ltd.
Although the authors are in no position to guarantee the claims made for any
particular reagent listed, the table is included to enable the reader to find what
reagents are available for the detection or determination of the commoner
metals, and some of the anions and miscellaneous substances that are included.

The figures in the table refer to the literature items in the list that follows the
table. Italic figures indicate that the methods referred to are qualitative only.
Some of the reagents and methods are sufficiently well established to appear in
standard text books and some of these are cited where they seem appropriate.
Apart from those that appear in the list, mention should also be made of six
further works that deal particularly with these methods, namely;—

NoelL Allport and John E. Brocksopp (1963). Colorimetric Analysis. Vol. 2.
Chapman and Hall.

E. B. Sandell, 3rd edn. (1959). Colorimetric Determination of Traces of Metals.
Interscience Publishers Inc.

G. Chariot (1964). Colorimetric Determination of Elements. Elsevier.

D. F. Boltz (1958). Colorimetric Determination of Non-metals. Interscience
Publishers Inc.

C. A., C. T. and F. D. Snell. Colorimetric Methods of Analysis. 6 vols. 1 948-1962.

D. van Nostrand.

R. E. Stanton (1966). Rapid Methods of Trace Analysis. London ; Edward Arnold
(Publishers) Ltd.

For the more recently introduced reagents or methods, reference is made to
papers in the journal literature. Only one paper is generally given for each
method and this has often been selected from many, either because it appears the
most important or because it is the most recent and provides, at the same time, a
key to previous writings on the subject.

871

i I III I e I S I 1 I I i I

^ ^ O «?a Oi 3 3 3 rt, 0 -fl;C.—

OSe4C6Ci5c/jc/jt/JC/3t/3c/ic/JW5H'r’HHr'

c W e y
3 *- .5 .13
> >« N N

Alizarin fluorine blue
Atorui fluorine blue
lanihanum complex

preparation

Alizarin reds

Barium chloranilate
Benzoin a-oxime

BerylloDll

42*-Bipyridyl

2.2»Biquinol>i

IHi

■■■BSaSSBSaSSa SS SSSSBSSSBSSMSB M^i

Call

Canniae
QJoratuIic acid
2-C[i!oro-4-nitro-
beazenediazomum
naphthalene*

^ 2-aulphoDate
Wimamoyl^pheoyl-
nydroxlaminf
Qeve’aacid
Pnpfcrron

- ^^PP^rill} ethyl acctoacelatc
Curciiniin

Jlaeiii.3.4-dilhli.»

gJl'l'Ml'Plitnylaatic

jS5JW:3Mthiol)

^‘^’'•Dibcnz^ilhio:

_ oxy ade

'^l^obcazyl-idane-

^'Jlbjlanunoniutn
-^^>!ithiocarb^te
• ■ d’?J^*’J**/^Pbenylcne-

^tPuno)ethane

mmmmammrn

.2 O S -S

8 E .§ g
= E c e
< < < <

I I I I -C ^ 2 1 a 1 ° I la

Suoi3(3uod3(3'EoocS

” 'S- 1 5

1*55

Dihydroxytartanc acid

Diroercaptothiadiazolc

Dimethylglyoxime

2,9'Dimcihyl-

1,10-phcoaothroIine
4,4'*Dinitrodiphcnj I-
carbazide

1 ,5>DiphenyIcarbazide
Dipheoylcarbazone
A^^-Diphcnylhydrarine
hydrochloride

4.7- Diphcnyl*
MQ-phenanthrolme

4.7- Diphcnyl-

M O^phenasthrolioedi-
sulphoaic acid dtsodium

salt

1,3-DipheayIpropane-

1,3-aioae

Dithio-oxamide

Dilhizone

Eriochrome cyanine R
Fonnaldoxime
, hydrochloride
. FuriJ a-dioairac
I' •

Hacmatoxylin
' Hexanitrodipbenylamioe
2-(HydroxymCTCury)-
benzoic acid
S-Hydxoiyquinaldioe
8-Hydroxyquinolinc
Lanthanum cbloraniJate
Magnesium blue
Magnesium uranyl acetate
Magnesons I and II
Mandelic aad
Mercaptoacetic add

2- Mcn:aptobcn2olhia2ole
Mcrcury(II) chloranilate

3- Melhoxyniirosopbeaol
a-Mcthoxyphcnylacclic

acid

Mctbylfluorone
Mordant red 74
Morin

NaphlhaJhydroxamic add
Niclccl uranyl acetate
Nioxime
Nitron

4- ;>-NilroiAcnyla20-
ordnol

1 -Nitroso-2-naphthol

mmamm

1 1 1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1 1

1 1 1 1 1 1

I I I I M I III! M I >1 I I I I I I I M M I I I

M I M I I I I I I I I I I I I I I I M I M I I I

HI I I
I I HI

i.sj 1

1 1591

1 1 1

1 1
|6J|59

l«l

i 1

1 60l 22l

1331 1

1 1

1 1 {

I ! 1

1 i

1 i *173

1 1

49!

1 1

1 1 >

I I HI
I I HI
H8| I I

MIN
MIN
I I I I I
Mill

I I I I I I I I I I I M M

2,9-D)jn«hyl-
4, 7-diphenyl-
I.IO-pbeaanthroline

Dimeibyl^yoxime

2,9-Dimethyl-

UO-pheoanthroline

1,5-DipheayIcarbazide

DiphenylcarbazoDe

Enochromecvamne R
Foncaldoumc
‘ hydrochloride
Ftini g-dioiimg

^cifl ’

Haematoxvlin

Hemutrodiphcnylamine

^•{Hydroxymcrcury)-

wnzoicacid

LHydroxvQuinalHiTi^

quinoline

jr^hanumchlQranihtff

I « i .1 i ^ I I

3 I s I 1 1 § §■ •§

.2 -te S *«• h ^ ^

ZZiZOCuA.04£liCU

I i 1 1 1 I E I s

11 1111 111 !

S e c S 5

S w S .5 s .5

.2 .2 T3 3 c

c CO n 3 o

— narscisyio
.2 .ts 3 »- « s .5 .3

HHH3>>-NN

mmmmmams

inHMi

■ ■■■■■ ■■■■■■■■■■■■■■■■■■■

lllllllllllllllllllllllll

MmMcSh

SHJ^^cadd~

~“ 2 Pi* 2 ! 20 toolc

MoEa 574

Molin ~ —

. ^loximc

Jfitron '

^ ^il

■■■■■■■■■■■■■■■■■■■■ ■■■■■■■■■■■■■■I

nprilMI

■■ ■B 8 BBBBBBB B BBBBBB Biiiha aMBia ^BBiBBBiB|
Bi EaaaSSSS8 BSgSgMSSggS8M aaaMa!EBBE BSigr'
■■■■■"BggggggggggggBBBMBBBBBBWWI

1^1

miiiiiiiiiiiiiiiiiiiiiue

— ■ ■“TBMBBBL.™

BHm 8 aaig 8 iKaH:a:MaEaBsii

■ aaMMMMBMMMMMMM pM aMMMMaaaMaBMa aaggai

iMgiBBMaBaaBaMBaaaBaaaBaBgBBEaaEB

875

876

Molybdenum

PotassM dibenz) Idithio-
caibamaie

H2-P)ridylazo)-

2-naphtbol

4-(2^d>’lazo)rcsorcinol

2-{2-Pyrid)l)-

benzimidazole

Su^er

diethylduhiocarbamate

Sodium

^ <b't[byldithiQcarbamate
^lumrhodizoaate

SMtis

Mbaio
Tannic acM
j j^ 6 '. 2 ''-Terpvrid\l

^nuoro^one

’Huourca ' ^

Thi crone
Thofin~

liroa

Banydiow

lESIlEI

8H BWg8aBliWaHS8MiB^81 1Ba

gKa MMBaaaBgBBaMaaBBg

miaaBanB

^MHW—mormwBBmmm

■■■■■■

It can be used to prepare a solution of toluene-3,4-dithiol and for many purposes it is added as the

877

List of literature references. Key to the table on the preceding pages

1 . Organic Reagents for Metals and other
Reagent Monographs, Vol. 1 , 5th Edn.,
1955. Hopkin & Williams.

2. Organic Reagents for Metals and for
Certain Radicals, Vol. II, 1964. Hopkins &
Williams.

3. /Ina/.C/iem.,33, 1128(1961).

4. /l«a/>>sr,87,703(1962).

5. Anal. Chim. Acta, 26, 487 (1962).

6. Anal. Chim. Acta, 27, 153(1962).

7 . Anal. Chem., 34, 209 (1962).

8. Anal. Chim. Acta, 27, 591 (1962).

9. Anal. Chim.Acta,27,i2\ (1962).

10. ra/an/a, 9, 987 (1962).

11. Anal. Chem., 35, 149 (1963).

12. Ind. Eng. Chem. (Anal.), IS, 57 (1943);
J.S.C.I., 62, 187(1943).

13. /laa/)W,73,395(1948).

14. Analyst,^!, 197(1962).

15. Anal. Chim. Acta, 13, 142 (1955).

16. /l«a/yit,87,880(1962).

17. /I«a/>’«,80,901 (1955).

18. /I«a/. C/i«n.,29,281 (1957).

19. Anatyst,94,262(m9).

20. Mikrochim. Acta,29(\962).

21. Anal. Chim. Acta, 26, 326 (1962).

22. Anal. Chem., 34, 571 (1962).

23. Anal.Abstr.,4,m%(\9S7).

24. Anal. Chun. Acta, 15,21, 102(1956).

25. J. Indian Chem. Sac., 28, 89 (1951); Chem.
Abstr., 45,^294(1951).

26. Anal. Abstr., 4, 3870 (1957).

27. ,4mi/./ltor.,4,3858(1957).

28. Anal. Chun. Acta, 19, 202 (1958).

29. J. Proc. Austral. Chem. Inst., 3, 184
(1926); Brit. Abstr., A. 1. 1082 (1936).

30. Anal. Chun. Acta, 19, 377 (1958).

31. Anal. Chim. Acta, 18, 546 (1958).

32. Z. anal. Chem., 140, 245 (1953); Anal.
Abstr., 1,659(1954).

33. A. I. Vogel (1961). Text-Book of
Quantitative Inorganic Analysis. 3rd edn.
Longmans.

34. J. Res. Nat. Bur. Stand., 33, 307 (1944);
Brit. Abstr., C, 91 (1945).

35. Ann. Chim., Roma, 43, 730 (1953) ; Anal.
Abstr.,\,47\ (1954).

36. Anal. Chem., 26, 883 (1954).

37. Anal. Chem., 34, 94 (1962).

38. ,4/ia/rs/, 82, 177(1957).

39. Analyst, 85, 276 (I960).

40. Anal. Chim. Acta, 25, 348 (1961).

41. Anal. Chem., 23, 653 (1951).

42. J. Inst. Petroleum Tech., 19, 845 (1933);
Brit. Abstr., B, 1088(1933).

43. F. Feigl (1958). Spot Tests in Inorganic
Analysis. 5th edn. Elsevier.

44. (a) Ind. Eng. Chem. (Anal.), 13, 603
(1941). (b) Anal. Abstr., 4, 2935 (1957).

45. Talanta, a, 579 (1961).

46. ra/aata, 5, 231 (1960).

47. /lna/>'st,85,823(1960).

48. Anal. Chem.,22, 1281 (1950).

49. A Handbook of Colorimetric Chemical
Analytical Methods. 6th edn. The
Tintometer Ltd.

50. A. 1. Vogel (1951). Text-Book of
Quantitative Inorganic Analysis. 2nd edn.
Longmans.

51. Anal. Chem.,32, 1337(1960).

52. Anal. Chem., 27, 1932 (1955).

53. ‘AnalaR’ S tandards for Laboratory
Chemicals. 5th edn., 1957. (Hopkin &
Williams and The British Drug Houses
Ltd), p. 71.

54. Analyst,95, 121 (1970).

55. (a) Anal. Chem., 25, 808 (1953). (b)
Analyst, 80, 768 (1955).

56. Brit. Abstr., A.1.215 (1942).

57. Anal. Chim. Acta, 19, 18(1958).

58. Anal. Chim. Acta, 30, 176(1964).

59. (a) Anal. Chem., 24, 1033 (1952). (b) Anal.
Abstr., 4,2135 (1957).

60. R. Belcher and C. L. Wilson (1956). New
Methods of Analytical Chemistry.
Chapman & Hall.

61. (a) Anal. Abstr., 6, 147, 916 (1959). (b)
Anal. Chim. Acta, 19, 576 (1958).

62. Anal. Abstr., 6, 869 (1959).

63. Analyst (abstr), 70, 189 (1945).

64. Anal. Chim. Acta, 19, 372 (1958).

65. Science, 125, 1042 (1957); Anal. Abstr., 5,
2560(1958).

66. Z. anal. Chem., 140, 252 (1953); Anal.
Abstr., 1,658 (1954).

67. Anal. Abstr. ,4, 2125 (1957).

68. Anal. Chim. Acta, 16, 121 (1957).

69. Anal. Abstr., 4, 1494 (1957).

70. Anal. Abstr., 4, 289 (1957).

71. Analyst,82, 620 (1957).

72. Anal. Chim. Acta, 22, 223 (1960).

73. Anal. Chem., 23, 514 (1951).

74. Anal. Abstr., 4, 286 (1957):

75. Anal. Chim. Acta, 12, 218 (1955).

76. Anal. Abstr., 4, 1444(1957).

77. Z. anal. Chem., 102, 24 (1935).

78. Z. anal. Chem., 102, 108 (1935).

79. Z. anal. Chem., 104, 88 (1936).

80. (a) Z. anal. Chem., 162, 96 (1958); Anal.
Abstr., 6, 944(1959). (b)Anal. Chim. Acta,
20,379(1959).

81. ,4na/.C/iem., 32, 1117(1960).

82. Z. anal. Chem., 171, 241 (1959); Anal.
Abstr.,7,2125(1960).

83. Analyst,89,707 (1964).

84. Ind. Eng. Chem. (Anal.), 12, 229 (1940).

85. Anal. Chem., 33, 1671 (1961).

86. Talanta, 9, 749 (1961).

87. Anal. Abstr.,3, 1244 (1956).

88. Anal. Chim. Acta,3, 481 (1949).

89. /lHa/yi/,85,889(1960).

90. Zna/yst, 58, 667 (1933).

91. Ind. Eng. Chem. (Anal.), 16, 222 (1944).

92. (a) Analyst, 56, 245 (1931). (b) Anal.

Abstr. ,3, 227 1(1956).

93. Anal. Chem., 28, 1728 (1956).

878

94. C/icm.v4ter.,45,69(1951). , .

95. Anal. Chim. Acta, 20, 340 (1959).

96. Anal. Chim. Acta, 20, 332 (1959).

97 Bull. Soc. chim. Belg., 54, 1 86 ( 1 945) ; Brit.

’. Abstr.,C.2{m]). ■ , ■

98. J. Chem. Soc. Japan, 66, 37 (1945) ; Chem.

,4Ai/r., 43, 1682(1949). .

99. (a) J. Indian Chem. Soc., 21, 1 19 (1944) ;

Analyst Abstr., 69, 383 (1944). (b) J. Indian
C/iem. 5oe., 21, 187,188 (1944); Br/4. .

Abstr.,C,92{m5).

100. ZavodLab., 11,254(1945); Chem. Abstr.,
40,1418(1946).

101. Anal. Chim. Acta, 21, 58 (1959).
101/1/10/^76,485(1951).

103. Analyst,16,m(l95l).

104. Analyst,19,5Ai{\954).

105. Anal. Chem.,26,2n (.1954).

106. Anal. Abstr., 6, 186, 187, 188(1959).

107. Anal. Chem., 35, 33 (1963).

108. Anal. Chem., 31, 1783 (1959).

109. Anal. Chim. Acta, 13, 154 (1955).

110. Anal. Chim. Acta, 13, 159 (1955).

111. Ind. Eng. Chem. (Anal.), 12, 663 (1940).

112. Ind. Eng. Chem. (Anal.), 14, 359 (1942).

113. Talanta,%, 612(1961).

114. Brit. Abstr.,C,39 (1944).

115. ro/a/ifa.8,293(1961).

116. Anal. Chem.,33, 421 (1961).

117. Anal.Abstr.,6,21l(\959).

118. Metallurgla, 44, 207 (1951); Brit. Abstr.,
C, 134(1952).

119. Anal. Chem.,22,9n (1950).

120. Anal. Chim. Acta, 23, 351 (1960).

121. Anal. Chim. Acta, 16, 62 (1957).

122. Anal. Chun. Acta, 13, 409 (1955).

123. Anal. Chim. Acta, 24, 294 (1961).

124. Anal. Abstr., 5, 1530(1958).

125. Anal. Chim. Acta, 2, 254 (1948) ; see also
ref. 60.

126. Analyst,iO, 220 (1955).

127. Analyst, 74, 505 (1949); 75, 555 (1950).

128. Ind. Eng. Chem. (Anal.), 16, 598 (1944).

129. Anal. Chim. Acta, 3, 324 (1949).

130. Analyst,10, \24(\945).

131. Mikrochim. Acta, 57t (1961).

132. Talanta, 2, 266 (1959).

133. Ta/onra, 8, 203 (1961).

134. Talanta,2,2\ (\959).

135. Talanta,2,152(\96\).

136. Anal. Chem.,23,445 (1961).

137. Anal. Chem.,31, 1102(1959).

138. Anal. Chem., 33, 239 (1961).

139. Anal. Chem., 33, 125 (1961).

140. 2. anal. Chem., 178, 352 (1961) ; Anal.
Abstr.,2, 3225(1961).

141. Anal. Chem., 26 , 1968 (1954).

142. .4/ia/>’of, 88, 380 (1963).

143. Talanta, 10, 1013 (1963).

144. Ta/o/i/a, 9, 761 (1962).

145. Ta/a/i/fl, 11, 621(1964).

146. Anal. Chim.'Acta,9,S6 (1952).

147. Anal. Abstr., 4, 1424 (1957).

148. Angew. Chem., 64, 608 (1952); Brit. Abstr.,
C, 246 (1953).

149. 83, 396(1958).

150. Anal. Chem., 25, 1 125 (1953).

151. Analyst, 58, 667 (1933); see also ref. 150.

152. Anal. Chem., 33, 1933 (1961).

153. /4/ifl/./4tor., 9, 106(1962). .

154. 'Anal. Abstr.,9^52 (1962.).

155. 83, 314 (1958).

156. - Analyst,84, 16(1959).

157. Anal. Chem., 29, 1187 (1957) (for
chloride);

158. Anal. Chim. Acta,22, 419 (I960).

159. Anal. Chim. Acta, 22, 413 (1960).

160. Tfl/anra, 7, 163(1961).

161. Zavod. Lab., 27, 803 (1961); W. Abstr.,

9, 558 (1962).

162. Analyst,S6,401 (1961).

163. Tfl/an/a, 4, 126(1960).

164. Talanta, 4, 244 (1960).

165. ra/a/i(fl, 11, 1 (1964).

166. Anal. Chem.,30, 44 (\952).

167. Anal. Chim. Acta, 21, 210 (1959).

168. Talanta,3,95(\959).

169. Anal. Chim. Acta,2A,\61 (\96\).

170. (a) Heh. Chim. Acta, 31, 320 (1948). (b)
Analyst, 74, 274 (1949).

171. 83, 516(1958).

172. Anal. Chim. Acta, 23, 175 (1960).

173. Anal. Chim. Acta, 23, 565 (1960).

174. Anal. Chim. Acta, 20, 26 (1959).

175. Anal. Chim. Acta, 23, 434 (1960).

176. /4na/. C/ie/7i., 32, 1350(1960).

177. Anal. Chim. Acta, 23, 538 (1960).

178. Anal. Chem., 32, 1083 (1960).

179. Ta/onM, 8,453(1961).

180. Analyst, 86, 543 (1961).

181. Anal. Chem., 31, 1985 (1959).

182. Anal. Chim. Acta,26, 522 (1962).

183. Zavod. Lab., 2B, 1283 (1957); C. A., 53,
11 R7/.nO‘iQV

184. Anal. Chim. Acta, 45, 341 (1969).

185. Mikrochim. Acta, 5\2(\96\).

186. Organic Chemical Reagents. Monograph
No. 70. 1967. Hopkin & Williams.

187. Organic Chemical Reagents. Monograph
No. 71 . 1967. Hopkin & Williams.

188. Analyst, 91,222 (1966).

189. J. Amer. Water Wks. Assocn., 49, 873
(1957); C. A., 51, 15045c (1957) (for
chlorine).

190. Monograph No. 74. 1969. Hopkin &
Williams.

191. Anal. Chim. Acta, 47, 151 (1969).

192. ‘AnalaR’ Standards for Laboratory
Chemicals. 6th edn., 1967 (‘AnalaR’
Standards Ltd), p. 617.

193. Monograph No. 75. 1969. Hopkin &
Williams.

879

APPENDIX III SPECIFIC GRAVITIES OF ACIDS AT 20 °C

(Specific gravities and percentages by weight are based on weights in vacup and
the percentage by weight refers to the formnla given)

Per cent
by weight

Specific gravity

H 2 SO 4

HNO 3

CHjCOOH

H 3 PO 4

HCl

1.0051

1.0036

0.9996

1.0038

1.0032

1.0118

1.0012

1.0092

1.0082

1.0184

1.0025

—

1.0201

1.0040

1.0200

1.0181

1.0317

1.0256

1.0055

—

1.0661

1.0543

1.0125

1.0532

1.0474

■IneH

1.0195

—

1.0209

1.0884

1.0776

1.1394

1.1150

1.0263

1.1134

1.0980

1.1404

1.0313

1.1395

1.1187

1.1783

1.1469

1.0326

—

1.1862

1.1534

1.0338

1.1529

1.1290

1.2185

1.1800

1.0384

1.1805

1.1493

1.2515

1.2071

1.0428

—

1.I69I

1.2599

1.2140

1.0438

1.216

—

1.2684

1.2205

1.0449

—

1.1789

1.3028

1.2463

1.0488

1.254

1.1980

1.3476

1.2783

1.0534

1.293

1.3951

1.3100

1.0575

1.335

1.4453

1.3393

1.0611

1.379

1.4983

1.3667

1.0642

1.426

1.5533

1.3913

1.0666

1.475

1.6105

1.4134

1.0685

1.526

1.6692

1.4337

1.0696

1.579

1.7272

1.4521

1.0700

1.633

1.7786

1.4686

1.0689

1.689

1.8144

1.4826

1.0661

1.746

1.8240

1.4873

1.0643

1.770

1.8279

1.4892

1.0632

—

1.8312

1.4912

1.0619

1.794

1.8337

1.4932

1.0605

1.8355

1.4952

1.0588

1.819

—

1.8364

1.4974

1.0570

—

1.8361

1.5008

1.0549

1.844

1.8342

1.5056

1.0524

1.8305

1.5129

1.0498

1.870

880

appendix IV SPECIFIC GRAVITIES OF ALKALINE SOLUTIONS
AT20°C

(Specific gravities and percentages by weight are based on weights in vacuo and
the percentage by w eight refers to the formula given)

Percent Specific gravity Percent Specific gravity

^ NaOH Nh" KOH NaOH NH3

1.0083

1.0095

0.9939

1.0175

1.0207

0.9895

1.0267

1.0318

—

1.0359

1.0428

0.9811

1.0452

1.0538

—

1.0544

1.0648

0.9730

1.0637

1.0758

—

1.0730

1.0869

0.9651

1.0824

1.0979

—

1.0918

1.1089

0.9575

1.1013

—

1.1108

1.1309

0.9501

1.1203

—

1.1299

1.1530

0.9430

1.1396

—

1.1493

1.1751

0.9362

1.1590

—

1.1688

1.1972

0.9295

1.1786

—

1.1884

1.2191

0.9229

1.1984

—

1.2083

1.2411

0.9164

1.2184

—

1.2285

1.2629

0.9101

1.2387

—

1.2489

1.2848

0.9040

1.2592

—

1.2695

1.3064

0.8980

1.2800

—

1.2905

1.3279

0.8920

1.3010

—

1.3117

1.3490

—

1.3224

—

1.3331

1.3696

—

1.3440

—

1.3549

1.3900

•

1.3659

1.3769

1.4101 .

1.3879

—

1.3991

1.4300

—

1.4103

—

1.4215

1.4494

—

1.4329

—

1.4443

1.4685

—

1.4558

—

1.4673

1.4873

—

1.4790

—

1.4907

1.5065

—

1.5025

—

1.5143

1.5253

—

APPENDIX V

DATA ON THE STRENGTH OF AQUEOUS SOLUTIONS
OF THE COMMON ACIDS AND OF AQUEOUS
AMMONIA

Reagent

Approximate

Percent Specific Normality
by weight gravity

Vol. required
to make 1 dm^
of approx. N
solution (cm^)

Hydrochloric acid

Nitric acid

Sulphuric acid

Perchloric acid

Hydrofluoric acid

Phosphoric acid

Acetic acid

99.5

Aqueous ammonia

27(NH3)

1.18

11.3

1.42

16.0

1.84

36.0

1.66

11.6

1.15

26.5

1.69

41.1

1.05

17.4

0.90

14.3

881

appendix VIII SOURCES OF ANALYSED SAMPLES

Throughout this book the use of a number of standard analytical samples is
recommended in order that practical experience may be gamed on substances of

known composition. . ,

In the United Kingdom such samples (sometimes known as Ridsdale s
Samples), suitable for metallurgical, chemical, and spectrographic analysis, are

Tk N^onal Bureau of Analysed Standards Ltd, Newham Hall,
Middlesborough, Yorkshire, England, from whom a detailed list is available
In the United States of America a similar wide range of standards, listea in

NBS Special Publication 260, 1975-6 Edition, can be obtained from .

US Department of Commerce, National Bureau of Standards, Washington
DC 20234 USA.

Elements and compounds of high purity and known composition are
marketed by Johnson, Matthey. Each batch has been subjected to spec roscopi
analysis and a detailed laboratory report accompanies each bate o ma e .
These are listed in booklet No. 1760, ‘Spectrographically Standardised

Substances’, and are supplied by; i

Johnson, Matthey Chemicals Ltd, Hatton Garden, London, . ng a .

An additional source of non-ferrous metal standards suitable or emission
spectroscopy and as chemical standards is the:

British Non-Ferrous Metals Technology Centre, Grove Laboratones,
Denchworth Road, Wantage, Oxfordshire, England.

885

appendix Vill SOURCES OF ANALYSED SAMPLES

Hioughout this book the use of a number of standard analytical samples is
recommended in order that practical experience may be gained on substances of

known composition. . , . ■

In the United Kingdom such samples (sometimes known as Ridsdale’s
Samples), suitable for metallurgical, chemical, and spectrographic analysis, are
supplied by:

The National Bureau of Analysed Standards Ltd, Newham Hall,
Middlesborough, Yorkshire, England, from whom a detailed list is available.

In the United States of America a similar wide range of standards, listed in
NBS Special Publication 260, 1975-6 Edition, can be obtained from:

US Department of Commerce, National Bureau of Standards, Washington
DC, 20234, USA.

Elements and compounds of high purity and known composition are
marketed by Johnson, Matthey. Each batch has been subjected to spectroscopic
analysis and a detailed laboratory report accompanies each batch of material.
These are listed in booklet No. 1760, ‘Spectrographically Standardised
Substances’, and are supplied by:

Johnson, Matthey Chemicals Ltd, Hatton Garden, London, ECl, England.

An additional source of non-ferrous metal standards suitable for emission
spectroscopy and as chemical standards is the :

British Non-Ferrous Metals Technology Centre, Grove Laboratories,
Denchworth Road, Wantage, Oxfordshire, England.

885

APPENDIX IX BUFFER SOLUTIONS AND SECONDARY pH
STANDARDS

The British standard for the pH scale is a 0.05M solution of potassium hydrogen-
phthalate (British Standard No. 1647: 1950 and 1961) which has a pH of 4.001 at
20 °C. The values at various other temperatures are collected belovv.

British pH standard: O.OSAf-potassium hydrogenphthalate

Temp., °C

4.011

4.020

4.105

'4.005

4.031

4.121

4.001

4.045

4.140

4.000

4.061

4.161

4.005

4.080

4.185,

4.011

4.091

4.211

Subsidiary pH standards at 25 °C include:

0.05Af-HCl + 0.09A/-Ka 2.07

0. 1 M-potassium tetroxalate 1 .48

0. 1 Af-potassium dihydrogen citrate 3.72

O.lAZ-acetic acid+0.1A/-sodium acetate 4.64

0.01 A/-acetic acid + 0.01 A/-sodium acetate 4,70

0.01A/-KHjPO., + 0.01A/-NajHP04 6.85

0.05A/-borax 9.18

0.025A/-NaHC03 + 0.025 AZ-Na^COj 1 0.00

0.01Af-Na3PO^ 11.72

The following table covering the pH range 2.6-12.0 (IS^C) is included as an
example of a universal buffer mixture.

A mixture of 6.008 g of A.R. citric acid, 3.893 g of A.R. potassium dihydrogen
phosphate, 1 .769 g of A.R. boric acid, and 5.266 g of pure diethylbarbituric acid
is dissolved in water and made up to 1 dm^. The pH values at 1 8 °C of mixtures of
100 cra^ of this solution with various volumes (X) of 0.2jW-sodium hydroxide
solution (free from carbonate) are tabulated below.

Ar(cm^)

ATCcm^)

Ar(cm’)

2.6

2.0

5.8

36.5

9.0

72.7

2.8

4.3

6.0

38.9

9.2

74.0

3.0

6.4

6.2

41.2

9.4

75.9

3.2

8.3

6.4

43.5

9.6

77.6

3.4

10.1

6.6

46.0

9.8

79.3

3.6

11.8

6.8

48.3

10.0

80.8

3.8

13.7

7.0

50.6

10.2

82.0

4.0

15.5

7.2

52.9

10.4

82.9

4.2

17.6

7.4

55.8

10.6

83.9

4.4

19.9

7.6

58.6

10.8

84.9

4.6

22.4

7.8

61.7

11.0

86.0

4.8

24.8

8.0

63.7

11.2

87.7

5.0

27.1

8,2

65.6

11.4

89.7

5.2

29.5

8.4

67.5

11.6

92.0

5.4

31.8

8.6

69.3

11.8

95.0

5,6

34.2

8.8

71.0

12.0

99.6

The National Bureau of Standards (NBS) pH standards (including notes on
the preparation of the buffer solutions) are given in Section XIV, 16, Table XIV, 3.

appendix X APPROXIMATE pH VALUES OF SOME COMMON
REAGENT SOLUTIONS AT ABOUT ROOM
TEMPERATURE

Add, benzoic
Add, boric
Add, dtric
Add, citric
Add, hydrochloric
Add, oxalic
Add, salicylic
Add, succinic
Add, tartaric
Ammonia, aqueous
Ammonium alum
Ammonium chloride

Ammonium dihydrogenphosphate
Ammonium oxalate

Diammonium hydrogenphosphate

Ammonium sulphate

Sodium tetraborate

Caldum hydroxide

Potassium acetate

Potassium alum

Potassium carbonate

Potassium dihydrogencitrate
Potassium dihydrogencitrate
Potassium hydrogencarbonate
Potassium dihydrogenphosphate
Potassium hydrogenoxalate
Sodium acetate
Sodium benzoate
Sodium carbonate
Sodium carbonate
Sodium hydrogencarbonate
Sodium hydrogensulphate
. dihydrogenphosphate
Sodium hydroxide
Ksodium hydrogen phosphate
insodmm phosphate

jvlphamic acid

Molarity

(Saturated)

2.8

0.1

5.3

0.1

2.1

0.01

2.6

0.1

1.1

0.1

1.3

' (Saturated)

2.4

0.1

2.7

0.1

2.0

0.1

11.3

0.05

4.6

0.1

4.6

0.1

4.0

0.1

6.4

0.1

7.9

0.1

5.5

0.1

9.2

(Saturated)

12.4

0.1

9.7

0.1

4.2

0.1

11.5

0.1

3.7

0.02

3.8

0.1

8.2

0.1

4.5

0.1

2.7

0.1

8.9

0.1

8.0

0.1

11.5

0.01

11.0

0.1

8.3

0.1

1.4

0.1

4.5

0.1

12.9

0.1

9.2

0.01

11.7

2.1

The pH values of 0 1 M

temperatures (°C) are; '

-HCl and of O.lM-NaOH

solutions at different

887

APPENDIX XI DISSOCIATION CONSTANTS OF SOME ACIDS IN
WATER AT 25 °C

Dissociation constants are expressed as pK^( - -log-^g)-

Add

pK,

Add

PK.

Aliphatic adds

Formic

Acetic

Propanoic

Butanoic

3-Methyl propanoic

Pentanoic

Fluoroacetic

CWoroacetic

Bromoacetic

lodoacetic

Cyanoacetic

Diethylacetic

Lactic

Pyruvic

Acrylic

■Vinylacelic

Tetrolic

trans-Crotonic

Furoic

Oxalic Ai

Malonic A,

Aromatic adds

Benzoic

Phenylacetic

Sulphanilic

Phenoxyacetic

Mandelic

1- Naphthoic

2- Naphthoic

1- Naphthylacetic

2- Naphthylacetic

3.75

Succinic

4.21

4.76

5.64

4.87

Glutaric

4.34

4.82

5.27

4.85

Adipic

4.43

4.84

5.28

2.58

Methylmalonic

3.07

2.86

5.87

2.90

Ethylmalonic

2.96

3.17

5.90

2.47

Dimethylmalonic

3.15

4,73

6.20

3.86

Diethylmalonic

2.15

2.49

7.47

4.26

Fumaric

3.02

4.34

4.38

2.65

Maleic

. 1.92

4.69

6.23

3.17

Tartaric

3.03

1.27

4.37

4.27

Citric

3.13

2.85

4.76

5.70

6.40

4.2!

4.31

2-Benzoylben2oic

3.54

3.23

Phthalic A,

.■ 2.95

3.17

A 2

5.41

3.41

cis-Cinnaroic

3.88

3.70

tro/ii-Cinnamic

4.44

4.16

Phenol

10.00

4.24

l-Nitroso-2-naphthoI

7.77

4.26

2-Nitroso- 1 -naphthol

7.38

ortho (2-)

mcta (3-)

para (4-)

Aromatic adds
Fluorobenzoic

3.27

3.86

4.14

Chlorobenzoic

2.94

3.83

3.98

Bromobenzoic

2.85

3.81

3.97 ,

lodobenzoic

2.86

3.85

3.93

Hydroxybenzoic

3.00

4.08

4.53

Methoxybenzoic

4.09

4.47

Nitrobenzoic

2.17

3.49

3.42

Aminobenzoic

4.98

4.79

4.92

Toluic

3.91

4.24

4.34

Chlorophenol

8.48

9.02

9.38

Nitrophenol

7.23

8.40

7.15

Methylphenol (cresol)

10.29

10.09

10.26

Methoxyphenol

9.98

9.65

10.21

888

APPENDIX Xil POTENTIALS OF THE COMMON REFERENCE
ELECTRODES

Electrode

Potential at 25 °C,
volts vs. N.H.E.

Hg/HgjCl^ (sat.), KCI (sat.) [S.C.E.]

+0.244

Hg/HgjClj (sat.), I.OM-KQ [N.C.E.]

+0.281

Hg/HgjCIj (sat.), O.IOA/-KC1

+ 0.336

Hg/HgjSO^ (sat.i KjSO^ (sat.)

+0.64

Hg/HgjSO^ (sat.), 0.05-4 A/-HjS04

+ 0.615

Ag/AgCI (sat.), KQ (sat.)

+0.199

AgjAgCX (sat.), 1.0A/-KC1

+0.227

Ag/Aga (sat.), O.lOAZ-KCl

+ 0.290

890

APPENDIX XIV TABLES OF ARC 'RAIES ULTIIWES' AND PERSISTENT
LINES FOR SPECTROGRAPHIC ANALYSIS*

The ‘Raies Ultimes’ and persistent lines marked in the enlargements of the
spectrograms of the R.U. powder (taken on a Hilger large quartz Littrow
spectrograph) are collected together in the table and are listed under the elements
to which they are due. The wavelengths are recorded in Angstroms, together with
their relative intensities as determined from the original enlargements.

The arc sensitivity of an element may be defined as the logarithm of the ratio of
the amount of diluent, such as the base used for the R.U. powder, to the amount
of the given element such that when the mixture is excited the strongest line of the
element will have an intensity 1. Thus a sensitivity of 6 corresponds to one part
per million, and a sensitivity of 1 to one part in 10.

The table also indicates the relative sensitivity, under arc excitation, of the
more important lines. Thus 'a' means the most sensitive and ‘b' the next in
sensitivity. The elements are grouped according to their absolute sensitivities in a
table at the end.

Barium contd.

Cadmium contd.

Cobalt contd.

Alumituum

4934.09

(51a

2288.02 (2) a

2407.25

(0)

3961.53

(7)a

4554.04 (15) a

3944.03

(5)0

~ Calcium

Copper

3092.71

(3)b

Beryllium

4455.89 (25)

3273.96 (15) a

3082.16

(3)b

3321.34

(2)

4434.96 (20)

3247.51 (20) a

- 3321.09

(2)

4425.44 (15)

Antimony

3131.07

(2)6

4318.65 (J2)

Fluorine

3267.50

(2)

3130.42

(3)b

CaF Band Head at A 5291

3232.50

2650.78

(0)

4283.01 (J2)

3029.81

(0)

2650.47

4226.73 (50)0

Gallium

2877.92

i-Oa

2494.73

(0)

3968.47 (40) a

4172.06

(5) a

2769.94

2494.56

(0)

3933.67 (40) a

2943.64

i4)

2598.06

(4) a

2348.61

(S)a

- 2874.24

(4)

2528.54

(4) a

~ Carbon

2311.47

(2) a

Bismuth

2478.57 (2) a

* Germaniom

(5) a

(2) a

3067,72

Arsem'c

2989.03

(2)6

Cesium

2754,59

(2) a

2898.71

(2)

2897.98

(2)6

4593.18 (l)a

2709.63

(2)0

2860.45

{3)b

- 4555.36 (3) a

2691.34

2780.20

{3)b

Boron

- 2651.58

2745.00

(2)

2497.73

(5)0

Chromium

2651.18

(2)0

2492.91

(2)

2496.78

(4) a

4289.72 (2)

2592.54

(1)

2456.53

(2)

- 4274.80 (2)

2381.18

(2)

Cadmium

4254.35 (2) a

Gold

2370.77

(2)

5085.82

3605.33 (2)

2675.95

(2)b

2369.67

(2)

4799.92

(1)

3593.49 (2)

2427.95

(2)0

2349.84

(4) a

3610.61

(3)b

3578.69 (3)

2288.12

{3)b

3467.66

(2)

Indium

- 3466.20

(2)6

Cobalt

4511.32

(6)0

Barium

3403.65

(2)

3453.51 (3) a

4101.77

(4)0

5535.55

(8) a

3261.06

(5) a

3405.12 (l)b

3258.56

(2)

* Reproduced by courtesy of the Research Laboratories of the General Electric Company Ltd,
Wembley, Middlesex.

Further details will be found in Booklet No. 1762, 'Sensitive Arc Lines of 50 elements including
the Use of R.U. Powder in Spectroscopic Analysis', available from Johnson, Matthey & Co. Ltd,
Hatton Garden, London, ECl. The R.U. Powder is available from this firm.

892

Iniim contd.

3256.09 (5)b
3039.36 (2)

Indium

3513.65

3220.78

3133.32

2639.71

2543.97

(J)6

(I)b

(!)a

(0)

Iron

3734.87

(2)

3719.94

(2)

3617.79

(2)

3581.20

(2)

3465.86

(2)

3440.61

(2)

3021.07

(2)

3020.64

(2)

2983.57

(0)

2973.24

(0)

2966.90

(0)

2719.03

(2)

2631,05

(0)

2599.40

(2)

2527.43

(0)

2522,85

(2)

2488.15

(2)

2483.27

(2) a

Lunthanum

4429.90

(2)

4333.73

(2)

4123.23

(2)

4086.71

(2)

3995.75

(2)

3949.11

(4) a

3337.49

(3)

lead

*57 .82 (103 a

3739.95

(2)

3683.47

(S)a

3639.58

(6) a

3572.73

(2)

2873.32

2833.07

(5) a

2823.19

(2)

2802.00

(-2)

2663,17

(3)

2614.18

(4)

2577.26

(2)

2476.38

(2)

2393.79

(2)

lithium
^’ 07.85 (20) a

Lithium contd.

Niobium contd. • Rhodium contd.

6103.64

(5)

3358.42

(2)

4602.86

(2)b

3232.61

(I)

Osmium

- 4260.85

(1)

Magnesium

. 3267.95

(2)

5183.62

(30)

3262.29

(1)

5172.70

(25)

3058.66

(2)

5167.34

(20)

3030.70

(1)

3838.26

(25)

3018.04

(1)

3832.31

(20)

2909.06

(3)0

3829.35

(15)

2838.63

(1)

3096.90

(30)

2488.55

(2)

3093.00

(15)

3091.08

(10)

Palladium

2852.13 (100) a

3634.70

(1)

2802.70

(30)

3609.55

(1)

2795.53

(30)

3516.94

(2)

~ 3481.15

(2)

Manganese

3460.77

(2)

4034.49

(3)0

3421.24

(3)

4033.07

(3)0

3404.58

(4)0

4030.76

(4)0

3242.70

(3)

im.n

(3)

2794.82

(1)

Phosphorus

2605.69

(I)

2554.93

(1)

2593.73

(1)

2553.28

(2)0

2576.10

(1)

2535.65

(2)0

- 2534.01

(1)

Mercury

2536.52

(2)0

Platiuum

- 3064.71

(2)0

Molybdenum

3042.64

(1)

3902.96

(4)0

2997.97

(1)

3864.11

(5)0

2929.79

(1)

3798.25

(5)0

2830.30

(1)

3447.12

(1)

2719.04

(1)

3358.12

(0)

2705.89

(0)

3208.83

(1)

2702.40

(1)

3193.97

(3)

2659.45

(2)0

3170.35

(3)

2650.86

(1)

3158.17

(2)

2628.03

(0)

3132.59

(3)

3112.12

(0)

Potassium

- 7698.98 i

(10)0

Nickel

7644.91 i

(10) a

3524.54

(2)b

4047.20

(5)b

3515.05

(0)

4044.14

(5)b

3446.26

(1)

3447.70

(1)

3414.77

(2) a

3446.72

(1)

3003.63

(0)

3002.49

(0)

Rhodium

- 4374.80

(l)b

Niobium

3692.36

(l)b

4123.81

(1)

3657.99

(0)

4100.92

(1)

3596.19

(0)

4079.73

(2)

3502.52

(1)

4058.94

(3)0

3434.89

(2)0

3396.85 (0)
3323.09 (0)

Ruthenium

3728.03

(2)

3726.93

(2)

3661.35

(1)

3634.92

( 1 )

3596.18

( 0 )

3593.02

(1)

3498.94

( 4)0

3436.74

(2)0

3428,63

(1)

3428.31

(1)

3417.35

( 0 )

2874.98

( 1 )

2735.72

( 1 )

2721.56

( 1 )

Rubidium

7947.60 (10) a

7800.23 (10) a

4215.56

(4)b

4201.85

(4)b

Scandium

4246.83

( 1 )

4023.69

( 2)0

4020.40

( 1)0

3911.81

( 2)0

3907.48

( 1)0

3642.79

( 0 )

3580.93

( 1 )

Silicon

3905.53

( 1 )

2881.58

( 8)0

2528.52

(4)

2524.12

(4)

2519.21

(4)

2516.12

( 5)0

2514.33

(4)

2506.90

(4)

2435.16

(2)

Silver

3382.89 (10) 0

3280.68 ( 10 ) 0

Sodium

5895.92 (30) 0

5889.95 (30) 0

5688.22

( 1 )

5682.66

( 0 )

3302.99

(4)b

3302.32

(4)b

893

Tia contd. Titanium contd.

Strontium

3034.12

(3)

3322.94 i

(0)

4607.33 (JO) a

3009.15

(I)

3241.99

(0)

4215.52

(6) a

2863.33

(4)0

3239.04

(7)

4077.71 (JO) a

2839.99

(4) a

3236.57

(7)

- 2706.51

(3)

3234.52

(7)

Tantalom

2661.25

(0)

3199.92

(7)

3318.84

(J)

2546.55

(7)

Tungsten

3311.16

(J)

2429.50

(2)

(0)b

3103.25

(0)

2354.85

(2)

4294.61

2714.67

(2) a

- 4008.75

(7) a

2656.61

(J)

Titam'um

2946.98

(0)

2653.27

(J)

4536.05

(7)

2944.40

(0)

2647.47

(J)

2646.22

(J)

4533.24

(7)

Vanadium

3998.64

(2)

4594.11

(0)

Thallium

3989.76

(J)

4460.29

(7)

5350.46

(7)a

3981.76

(0)

4408.51

(I)

3775.72

(6) a

3958.21

(J)

3529.43

(4)b

3956.34

(0)

4379.24

(2)

3519.24

(S)b

3653.50

(J)

4134.49

(0)

3229.75

(J)

3642.68

(J)

4132.02

(7)

2921.52

(7)

3635.46

(J)

4128.07

(2)

2918.32

(I)

3377.59

(J)

4099.80

(7)

2767.87

(3)

3372.80

(2) '

3185.40

(2)

2580.14

(J)

3371.45

(2)

3183.98

(2)

2379.69

(/)

3370.44

(7)

3183.41

(7)

3361.21

(2)

3110.71

(0)

Tin

3354.64

(J)

3102.30

(7)

3330.59

(2)

3349.41

(2)

3066.38

(7)

3262.33

(4) a

3349.04

(7)

3060.46

(7)

3175.02

(3)

3341.88

(i)

2924.64

(7)

Vanadium contd.
2924.03 (0)
2923.62 (0)
2908.82 (i)

Yttrium

4643.70

(0)

4374.94

(7)'

4142.84

{2)

4102.38

[3)0

3982.60

(0)

3774.33

(2)

3710.29

(2)

3620.94

(7)

3242.28

(-’)

3216.68

(7)

Zinc

6362.35 (20) b
4810.53 (SO) a
4722.16 (SO) a

4680.14 (30) a

3345.93 (30) b
3345.57 (30)b

3302.94 (2S)
3302.59 (25) ,
3282.33 (20) ,

Zirconium
3496.21 (2) a

3481.15 (0)
3438.23 (2) a
3391.98 (2) a

Arc Sensitivity between 5 and 6 :
Ag, Co, Cr, Cu, In, Li, Mg,

Na, Ni, Os, Pd, Pt, Rh, Ru.

Arc Sensitivity between 4 and 5 :

Al, Au, Ba, Be, Ca, Fe, Ga, Ge, Hg,
It, K, Mn, Mo, Pb, Sc, Sn, Sr. Ti, V.

Arc Sensitivity between 3 and 4 :
B, Bi, Cd, La, Rb, Sb,

Si, n, Y, Zn, Zr.

Arc Sensitivity between 2 and 3:
As, Cs, Nb, P, Ta, W.

894

appendix XV PERCENTAGE POINTS OF THE f-DISTRIBUTION

the percentage of the area under the two tails of the /-curve, and therefore gives the
* y that t will exceed the tabular entry in absolute value.

XVI and XVII have been derived, with the permission of the
metnka Trustees, from the corresponding tables in Biometrika Tables for
Vol. I, Third Edition (1966), and Biometrika Tables for

Cambridge

895

APPENDIX XVI /^-DISTRIBUTION

Probability <^2
level

A , (corresponding to greater mean square)

1 2 3 4, 5 6

OJO

39.9

49.5

53.6

55.8

57.2

58.2

58.9

59,4

59.9

60.2

61.2

63.3

0.05

161.4

199.5

215.7

224.6

230.2

234.0

236.8

238.9

240.5

241.9

246.0

254.3

0.01

4,052

4,999

5,403

5,625

5,764

5,859

5,928

5,981

6,023

6,056

6,157

6,366

OJO

8.53

9.00

9.16

9.24

9.29

9,33

9.35

9.37

9.38

9.39

9.42

9.49

0.05

18.5

19.0

19.2

19.3

19.4

19.5

O.OI

98.5

99.0

99.2

99,2

99.3

99.4

■ 99.4

99.4

99.5

0.10

5.54

5.46

5.39

5.34

5.31

5.28

5.27

5.25

5.24

5.23

5.20

5.13

0.05

10.1

9.55

9.28

9.12

9.01

8.94

8.89

8.85

8.81

8.79

8.70

8.53

0.01

34.1

30.8

29.5

28,7

28.2

27.9

27.7

27.5

27.3

27.2

26.9

26.1

0.10

4.54

4.32

4.19

4.11

4.05

4.01

3.98

3.95

3.94

3.92

3.87

3.76

0.05

7.71

6.94

6.59

6.39

6.26

6.16

6.09

6.04

6.00

5.96

5.86

5.63

0.01

21.2

13.0

16.7

15.0

15.S

15.2

15.0

14.8

14.7

14.5

14.2

13.5

0.10

4.06

3.78

3.62

3.52

3.45

3.40

3.37

3.34

3.32

3.30

3.24

3.10

0.05

6.61

5.79

5.41

5.19

5.05

4.95

4.88

4.82

4.77

4.74

4.62

4.36

0.01

16.3

13.3

12.1

11.4

1 1.0

10.7

10.5

10.3

10.2

10.1

9.72

9.02

0.10

3.78

3.46

3.29

3.18

3.11

3.05

3.01

2.98

2.96

2.94

2.87

172

0,05

5.99

5.14

4.76

4.53

4.39

4.28

4.21

4.15

4.10

4.06

3.94

3.67

0.01

13.7

10,9

9.78

9.15

8.75

8.47

8.26

8.10

7.98

7,87

7.56

6.88

0.10

3.59

3.26

3.07

2.96

2.88

Z83

2.78

2.75

2.72

2.70

2.63

2.47

0.05

5.59

4.74

4.35

4.12

3.97

3.87

3.79

3.73

3.68

3.64

3.51

3.23

0.01

12.2

9.55

8.45

7.85

7.46

7.19

6.99

6.84

6.72

6.62

6.31

5.65

0.10

3.46

3.11

2,92

2.81

2.73

2.67

2.62

2.59

2.56

2J4

2.46

129

0.05

5.32

4.46

4.07

3,84

3,69

3.58

3.50

3.44

3.39

3.35

3,22

193

0.01

11.3

8.65

7.59

7.01

6.63

6.37

6.18

6.03

5.91

5.81

5.52

4.86

0.10

3.36

3.01

2,81

2.59

2.61

2.55

2.51

2.47

2.44

2.42

2.34

116

0.05

5.12

4.26

3.86

3.63

3.48

3.37

3.29

3.23

3.18

3.14

3.01

2.71

0.01

10.6

8,02

6.99

6.42

6.05

5.80

5.61

5.47

5.35

5.26

4.96

4.31

0.10

3.29

2.92

2.73

2.61

2.52

2.46

2.41

2.38

2.35

2.32

2.24

106

0.05

4.96

4.10

3.71

3.48

3.33

3.22

3.14

3.07

3.02

2.98

2.85

154

0.01

10.0

7.56

6.55

5.99

5.64

5.39

5.20

5.06

4.94

4.85

4.56

3.91

0.10

3.18

2.81

2,61

2.48

2.39

2.33

2.28

2.24

121

119

110

1.90

0.05

4.75

3.89

3.49

3.26

3.11

3.00

2.91

2.85

180

2.75

162

2.30

0.01

9.33

6.93

5.95

5.41

5.06

4.82

4.64

4.50

4.39

4.30

4.01

3.36

0.10

3.07

2.70

2,49

2.36

2,27

2.21

2.16

2.12

2.09

2.06

1.97

1.76

0.05

4.54

3.68

3.29

3.06

2.90

2.79

2.71

2.64

2.59

2.54

2.40

2.07

0.01

8.68

6.36

5.42

4,89

4,56

4.32

4.14

4,00

3.89

3.80

3.52

187

0.10

3.05

2.67

2.46

2,33

2.24

2.18

2.13

2.09

2.06

2.03

1.94

1.72

0.05

4.49

3.63

3.24

3,01

2.85

2.74

2.66

2.59

154

2.49

135

2.01

0.01

8.53

6.23

5.29

4.77

4,44

4.20

4.03

3.89

3.78

3.69

3.41

2.75

0.10

2.93

2.54

2.33

2.19

2.10

2.04

1.98

1.94

1.91

1.88

1.78

1.53

0.05

4.26

3.40

3.01

2.78

2.62

2.51

2.42

2.36

130

125

111

1.73

0.01

7.82

5.61

4.72

4.22

3.90

3.67

3.50

.3.36

3.26

3.17

2.89

2.21

0.10

2.79

2.39

2.18

2.04

1.95

1.87

1.82

1.77

1.74

1.71

1.60

1,29

0.05

4,00

3.15

2.76

2.53

2.37

2.25

2.17

2.10

2.04

1.99

1.84

1.39

0.01

7.08

4.98

4,13

3.65

3.34

3.12

2.95

2,82

2.72

2.63

135

1.60

0.10

2.71

2.30

2.08

1.94

1.85

1.77

1.72

1.67

1.63

1.60

1.49

1.00

0.05

3.84

3.00

2.60

2.37

2.21

2.10

2.01

1.94

1.88

1.83

1.67

,1.00

0.01

6.63

4.61

3.78

3.32

3.02

2.80

2.64

2.51

141

2.32

2.04

1.00

896

appendix XVII PERCENTAGE POINTS OF THE -DISTRIBUTION

i 99 % 97 . 5 % 95 % 90 % 50 %

0.000

0.001

0.004

0.016

0.455

0.020

0.051

0.103

0.211

1.39

0.115

0.216

0.352

0.584

2.37

0.297

0,484

0.711

1.06

3.36

0.554

0.831

1.15

1.61

4;35

0.872

1.24

1.64

2.20

5.35

1.24

1.69

2.17

2.83

6.35

1.65

2.18

2.73

3.49

7.34

2.09

2.70

3.33

4.17

8.34

2.56

• 3.25

3.94

4.87

9.34

3.05

3.82

4.57

5.58

10.34

3.57

4.40

5.23

6.30

11.34

4.11

5.01

5.89

7.04

12.34

4.66

5.63

6.57

7.79

13.34

5.23

6.26

7.26

8.55

14.34

5,81

6.91

7.96

9.31

15.34

6.41

7.56

8.67

10.09

16.34

7.01

8.23

9.39

10.86

17.34

7.63

8.91

10.12

11.65

18.34

8.26

9.59

10.85

12.44

19.34

8.90

10.28

11.59

13.24

20.34

9.54

10.98

12.34

14.04

21.34

10.20

11.69

13.09

14.85

22.34

10.86

12.40

13.85

15.66

23.34

11.52

13.12

14.61

16.47

24,34

12.20

13.84

15.38

17.29

25.34

12.88

14.57

16.15

18.11

26.34

13.56

15.31

16.93

18.94

27.34

14.26

16.05

17.71

19.77

28.34

14.95

16.79

18.49

20,60

29.34

22.16

24.43

26.51

29.05

39.34

29.71

32.36

34.76

37.69

49.34

37.49

40.48

43.19

46.46

59.33

45.44

48.76

51.74

55.33

69.33

lOO

53.54

57.15

60.39

64.28

79.33

61.75

70.06

65.65

74.22

69.13

77.93

73.29

82.36

89.33

99.33

10 %

5 %

2 . 5 %

1 %

0 . 1 %

2.71

3,84

5.02

6.63

10.83

4.61

5.99

7.38

9.21

13.82

6.25

7.81

9.35

11.34

16.27

7.78

9.49

11.14

13.28

18.47

9.24

11.07

12.83

15.09

20.52

10.64

12.59

14.45

16.81

22.46

12.02

14.07

16.01

18.48

24.32

13.36

15.51

17.53

20.09

26.13

14.68

16.92

19.02

21.67

27.88

15.99

18.31

20.48

23.21

29.59

17.28

19.68

21.92

24.72

31.26

18.55

21.03

23.34

26.22

32.91

19.81

22.36

24.74

27.69

34.53

21.06

23.68

26.12

29.14

36.12

22.31

25.00

27.49

30.58

37.70

23.54

26.30

28.85

32.00

39.25

24.77

27.59

30.19

33.41

40.79

25.99

28.87

31.53

34.81

42.31

27.20

30.14

32.85

36,19

43.82

28.41

31.41

34.17

37.57

45.32

29.62

32.67

35.48

38.93

46.80

30.81

33.92

36.78

40.29

48.27

32.01

35.17

38.08

41.64

49.73

33.20

36.42

39.36

42.98

51.18

34.38

37.65

40.65

44.31

52.62

35.56

38.89

41.92

45.64

54.05

36.74

40.11

43.19

46.96

55.48

37.92

41.34

44.46

48.28

56.89

39.09

42.56

45.72

49.59

58.30

40.26

43.77

46.98

50.89

59.70

51.81

55.76

59.34

63.69

73.40

63.17

67.50

71.42

76.15

86.66

74.40

79.08

83.30

88.38

99.61

85.53

90.53

95.02

100.42

112.32

96.58

101.88

106.63

112.33

124.84

107.57

113.14

118.14

124.12

137.21

118.50

124.34

129.56

138.81

149.45

897

APPENDIX XVIll FOUR-FIGURE LOGARITHMS

Mean differences

1 2 3 4 5

7 8 9

0000

0043

0086

0128

0170

0212

0253

0294

0334

0374

17 21

25 29 33 37

0414

0453

0492

0531

0569

0607

0645

0682

0719

0755

19 23 26 30 34

0792

0828

0864

0899

0934

0969

1004

1038

1072

1106

17 21

24 28 31

1139

1173

1206

1239

1271

1303

1335

1367

1399

1430

19 23 26 29

1461

1492

1523

1553

1584

1614

1644

1673

1703

1732

18 21 24 27

1761

1790

1818

1847

1875

1903

1931

1959

1987

2014

17 20 22 25

2041

2068

2095

2122

2148

2175

2201

2227

2253

2279

18 21 24

2304

2330

2355

2380

2405

2430

2455

2480

2504 ■

2529

17 20 22

2553

2577

2601

2625

2648

2672

2695

2718

2742

2765

16 19 21

2788

2810

2833

2856

2878

2900

2923

2945

2967

2989

16 18 20

3010

3032

3054

3075

3096

3118

3139

3160

3181

3201

15 17 19

3222

3243

3263

3284

3304

3324

3345

3365

3385

3404

14 16 18

3424

3444

3464

3483

3502

3522

3541

3560

3579

3598

14 15 17

3617

3636

3655

3674

3692

3711

3729

3747

3766

3784

6 '

13 15 17

3802

3820

3838

3856

3874

3892

3909

3927

3945

3962

12 14 16

3979

3997

4014

4031

4048

4065

4082

4099

4116

4133

12 14 15

4150

4166

4183

4200

4216

4232

4249

4265

4281

4298

11 13 IS

4314

4330

4346

4362

4378

4393

4409

4425

4440

4456

11 13 14

4472

4487

4502

4518

4533

4548

4564

4579

4594

4609

■8

11 12 14

4624

4639

4654

4669

4683

4698

4713

4728

4742

4757

10 12 13

4771

4786

4800

4814

4829

4843

4857

4871

4886

4900

9 10 11 13

4914

4928

4942

4955

4969

4983

4997

5011

5024

5038

10 11 12

5051

5065

5079

5092

5105

5119

5132

5145

5159

5172 .

9 11 12

5185

5198

5211

5224

5237

5250

5263

5276

5289

5302

9 10 12

5315

5328

5340

5353

5366

5378

5391

5403

5416

5428

9 10 11

5441

5453

5465

5478

5490

5502

5514

5527

5539

5551

9 10 11

5563

5575

5587

5599

5611

5623

5635

5647

5658

5670

8 10 11

5682

5694

5705

5717

5729

5740

5752

5763

5775

5786

8 9 10

5798

5809

5821

5832

5843

5855

5866

5877

5888

5899

8 9 10

5911

5922

5933

5944

5955

5966

5977

5988

5999

6010

. 4

8 9 10

6021

6031

6042

6053

6064

6075

6085

6096

6107

6117

1 ,

8 9 10

6128

6138

6149

6160

6170

6180

6191

6201

6212

6222

7 8 9

6232

6243

6253

6263

6274

6284

6294

6304

6314

6325

7 8 9

6335

6345

6355

6365

6375

6385

6395

6405

6415

6425

7 8 9

6435

6444

6454

6464

6474

6484

6493

6503

6513

6522

7 8 9

6532

6542

6551

6561

6571

6580

6590

6599

6609

6618

7 8 9

6628

6637

6646

6656

6665

6675

6684

6693

6702

6712

7 7 8

6721

6730

6739

6749

6758

6767

6776

6785

6794

6803

6 7 8

6812

6821

6830

6839

6848

6857

6866

6875

6884

6893

6 7 8

6902

6911

6920

6928

6937

6946

6955

6964

6972

6981

6 7 8

6990

6998

7007

7016

7024

7033

7042

7050

7059

7067

6 7 8

7076

7084

7093

7101

7110

7118

7126

7135

7143

7152

6 7 8

7160

7168

7177

7185

7193

7202

7210

7218

7226

7235

6 7 7

7243

7251

7259

7267

7275

7284

7292

7300

7308

7316

6 6 7

7324

7332

7340

7348

7356

7364

7372

7380

7388

7396

6 6 7

7 8 9

898

appendix XVIII FOUR-FIGURE LOGARITHMS

Mean differences

1 2 3 4 5

7404

7412

7419

7427

7435

7443

7451

7459

7466

7474

7482

7490

7497

7505

7513

7520

7528

7536

7543

7551

7559

7566

7574

7582

7589

7597

7604

7612

7619

7627

7634

7642

7649

7657

7664

7672

7679

7686

7694

7701

7709

7716

7723

7731

7738

7745

7752

7760

7767

7774

7782

7789

7796

7803

7810

7818

7825

7832

7839

7846

7853

7860

7868

7875

7882

7889

7896

7903

7910

7917

7924

7931

7938

7945

7952

7959

7966

7973

7980

7987

7993

8000

8007

8014

8021

8028

8035

8041

8048

8055

8062

8069

8075

8082

8089

8096

8102

8109

8116

8122

8129

8136

8142

8149

8156

8162

8169

8176

8182

8189

8195

8202

8209

8215

8222

8228

8235

8241

8248

8254

8261

8267

8274

8280

8287

8293

8299

8306

8312

8319

8325

8331

8338

8344

8351

8357

8363

8370

8376

8382

8388

8395

8401

8407

8414

8420

8426

8432

8439

8445

8451

8457

8463

8470

8476

8482

8488

8494

8500

8506

8513

8519

8525

8531

8537

8543

8549

8555

8561

8567

8573

8579

8585

8591

8597

8603

8609

8615

8621

8627

8633

8639

8645

8651

8657

8663

8669

8675

8681

8686

8692

8698

8704

8710

8716

8722

8727

8733

8739

8745

8751

8756

8762

8768

8774

8779

8785

8791

8797

8802

8808

8814

8820

8825

8831

8837

8842

8848

8854

8859

8865

8871

8876

8882

8887

8893

8899

8904

8910

8915

8921

8927

8932

8938

8943

8949

8954

8960

8965

8971

8976

8982

8987

8993

8998

9004

9009

9015

9020

9025

9031

9036

9042

9047

9053

9058

9063

9069

9074

9079

9085

9090

9096

9101

9106

9112

9117

9122

9128

9133

9138

9143

9149

9154

9159

9165

9170

9175

9180

9186

9191

9196

9201

9206

9212

9217

9222

9227

9232

9238

9243

9248

9253

9258

9263

9269

9274

9279

9284

9289

9294

9299

9304

9309

9315

9320

9325

9330

9335

9340

9345

9350

9355

9360

9365

9370

9375

9380

9385

9390

9395

9400

9405

9410

9415

9420

9425

9430

9435

9440

9445

9450

9455

9460

9465

9469

9474

9479

9484

9489

9494

9499

9504

9509

9513

9518

9523

9528

9533

9538

9542

9547

9552

9557

9562

9566

9571

9576

9581

9586

9590

9595

9600

9605

9609

9614

9619

9624

9628

9633

963li

9643

9647

9652

9657

9661

9666

9671

9675

9680

9685

9689

9694

9699

9703

9708

9713

9717

9722

9727

9731

9736

9741

9745

9750

9754

9759

9763

9768

9773

9777

9782

9786

9791

9795

9800

9805

9809

9814

9818

9823

9827

9832

9836

9841

9845

9850

9854

9859

9863

9868

9872

9877

9881

9886

9890

9894

9899

9903

9908

9912

9956

9917

9921

9926

9930

9934

9939

9943

9948

9952

9961

9965

9969

9974

9978

9983

9987

9991

9996

—

899

INDEX

The following abbreviations are used :

aa = atomic absorption

= flame emission

sepn. = separation

am = araperometry

= fluorimetric

soln. = solution

eh = chromatographic

= gravimetric

stdn. = standardisation

cm = coulometric

= high frequency

temp. = temperature

cn = conductometric

= potentiometric

th = thermal

D = determination

prep.

= preparation

ti = titrimetric -

eg = electrogravimetric

= spectrophotometric

V = voltammetry

em = emission spectrographic

= solvent extraction

Absorbance: 696, 759, 764
Absorbancy: 697
Absorbents: 492
Absorption coefiBcient: 696, 697
Absorption curve, D. of: with a

spectrophotometer, of methyl red, 761
of potassium dichromate, 763
of potassium nitrate, 759
of potassium permanganate, 763
a.c. spark source: 793
Accelerator: 109
Accuracy: 7, 10, 121
absolute, 121
comparative, 122
in quantitative analysis, 121
Acetic acid: D. of strength of, (ti) 308
specific gravities of aqueous solns., 880, 881
Acetylacetone: 149
chelation complexes with, 150, 218
Acid-base indicators: 236
prep, of solns. of, 242
table of, 240

Acid-base titrations: 236, 296
theory of, 244

Acidimetry and alkalimetry: 224, 236
theory of, 244

Acids: Bronsted theory of, 15
common, strengths of, 881
dissociation constants of, 888
hard, 259

Acids—continued
ionisation of, 14
Lewis, 17, 21
polyprotic, 14, 22
prep, of standard solns., 296, 297
soft, 259

specific gravities of, 880
strengths of, 20, 307, 881
strong, 15

pK values in aqueous soln., 888
weak, 15

Acids, standardisation of, 298, 300; 385
Acids, titration of: by hydroxide ion (cm), 562
Activity: 17, 18
coefficient, 18
Adsorption: 409
Adsorption indicators: 284
applications of, 284, 338, 339
choice of, 285
table of, 287
theory of, 284
Agar

-potassium chloride bridge, 60S
-potassium nitrate bridge, 604
Ageing of precipitates: 440
Air bath: 90

Alizarin Yellow R: 240, 243
Alkaline solutions: specific gravities of, 881
Aluminium, D. of: as tris(acetylacetonato)
complex, (ch) 218

901

INDEX

Alaminiutn, D. of— continued
as oxide, (g) 435, 449
as oxinate, (g) 436, (ti) 393
by EDTA,(p)610,(ti)3I9
by Eriochrome Cyanine R, (s) 729
fiuorimetric method, (fu) 777
Aluminium, sepn. of: plus iron, (ch) 202
Amalgams: jee Liquid Amalgams
Amaranth: 387

Ammonia, D. of: in an ammonium salt, (ti) 312
with Nessler’s reagent, (s) 730
Ammonia solution, strength of: 88 1
Ammonium, D. of: as chloroplatinate, (g) 450
as tetraphenylborate, (g) 450
Ammonium cerium(III) nitrate: purification of,
365

Ammonium cerium(IV) sulphate: 365
Ammonium iron(III) sulphate indicator: 341
Ammonium magnesium phosphate, thermal
analysis of: 859

Ammonium molybdate reagent: prep, of, 500,
504, 732, 757

Ammonium thiocyanate soln. : prep, of, 34 1
stdn. of, 341
use of, 340

Ammonium vanadate: 757
Ampere: 515

Amperometric titrations: advantages of, 672,
biamperomctric, 685
cells for, 675

common types of titration curves, 674
determinations by, 676
theoretical considerations, 672
use of dropping mercury electrode, 675
use of rotating platinum micro-electrode,
682

with two indicator electrodes (dead stop end
point), 685
Amperostat : 649
Amplification methods: 125
Analysed samples and standards: list of sources,
885

Analysis: complete, 4
macro, 4
micro, 4
partial, 4
proximate, 4
qualitative, 3
quantitative, 3
semi-micro, 4
trace constituent, 4
Angstrom: 404
Anion exchangers: 165, 302
strongly basic, 169, 172
weakly basic, 169, 172

Anion exchanges— continued
see also Ion Exchange
Anode: 516, 639

Anodic stripping voltammetry; 633, 664
basic principles, 664
fundamental features, 665
instrumentation, 667
Anhydrone: 492
Anthranilic acid : 427
Antibumping granules: 313
Antimony, D. of: as antimony(III) pyrogallate,
(g)451

as trisulphide, (g) 450
by iodine, (cm) 557, (ti) 383
by potassium bromate, (am) 684, (ti) 392
by potassium iodate, (ti) 387
by potassium iodide, (s) 751
in presence of arsenic, (s) 769
Antimony electrode: use of, 572
Antipyrine : see Phenazone
Apparent indieator constant: 237, 239
Apparent stability constant ; 264
Arc: alternating current, 793
direct current, 792
sensitivities of elements, 796
Arsenates, D. of : (ti) 344
.Arsenic, D. of: as ammonium uranyl arsenate,
(g)452

as silver arsenate, (ti) 344
as trisulphide, (g) 45 1
by iodine, (am) 684, (ti) 383
by molybdenum blue method, (s) 732
by potassium bromate, (ti) 392
by potassium iodate, (ti) 387
by silver diethyl dithiocarbamate, (s) 734
in presence of antimony, (g) 452, (s) 769
Arsenic(IlI) oxide; as primary standard, 236
Arsenite, D. of: (am) 684
Arsonic acids: 427

Atomic absorption inhibition titrimetry: 843
Atomic absorption spectrophotometers: 814
single beam, 826

double beam, 827 , ,

Atomic absorption spectroscopy: 810, 813
data for common elements, 834
determinations, 834
non-flame techniques, 817
instrumentation, 814, 826, 827
interferences, 821
theory, 811

Atomic excitation, table of: 8 12
Atomic fluorescencespectroscopy: 810, 813, 830
instrumentation, 814
Atomic weights, table of: 870
Autodispensers: 78

902

INDEX

Automatic analysis : 1 53
Automatic potentiometric titrations: 594 . ■
instrumentation, 594 ,

Available chlorine: see Bleaching powder
AzoUtmia:241

Back e.m.f.: 517
Back titration : 266,' 3 i 9
Balance: analytical, free-swinging, 59
aperiodic, 62 . -

care and use, 68

constant load, 64 ... .

controlled release of, 64

electronic, 66 ■ '

maximum load of, 60

micro, 60

preweighing devices, 63
requirements of, 60
rider, 60
semimicro, 60
sensitivity of, 61
single-pan, 62
theory of, 59
top loading, 65
torsion, 67

Balancing method: 700, 705
Barium, D. of: as chromate, (g) 453, (ti) 362
as sulphate, (g) 452
byEDTA, (ti)3l9
Barium chloranilate: 758
Barium chloride, D. of water of hydration; (s)
432

Barium hydroxide, standard solution: 306
Barium sulphate: sepn. from supersaturated
soln., 409

Barrier-layer cells: 708
Baryta: see Barium hydroxide
Bases: Bronsted theory of, 15
dissociation constants of, 889
hard, 259
ionisation of, 14
Lewis, 1 7
soft, 259
strengths of, 20
strong, 15

pK values in aq. soln., 889
. titration of with hydrogen ion, (cm) 564
weak, 15

Basic acetate method : 4 1 9
Basic benzoate method: 419
Beakers: 84
tongs for, 91

Bearing metal, analysis of; (eg) 539
Beer-Lambert law: 695, 697, 773

Beer’s law: 695, 696, 729 , ,

application of, 697, 761
deviations from, 698 ; . ;

Benzoic acid: 305 .. ..

Benzoin-a-oxime: see Cupron
7V-Benzoyl-N-phenylhydroxylamine : 422, 484
Beryllium, D. of: as acetylacetone complex, (se)

153 . . -

as oxide, (g) 454

as pyrophosphate, (g) 455 , ' i . '

by4-nitrobenzene-azo-orcinol, (s)735 , -
by sodium hydroxide, (hf ) 630
Biamperometric titrations: see Amperometric
titrations

Bismuth, D. of: as oxyiodide, (g) 455
as pyrogallate, (g) 456
by EDTA, (li) 610, (am) 320 . -

by potassium iodide method, (s) 735
in lead, (s) 736

in presence of cadmium and lead, (ti) 333 .
Bismuth liquid amalgam: 397
Blank determination r 1 24
Bleaching powder, D. of available chlorine: 382
by arsenite method, (ti) 382
by iodometric method, (ti) 382
Boiling rods: 94
Boltzmann equation; 81 1
Borate, D. of: as nitron tetrafluoborate, (g) 491
by l,r-dianthrimide, (s) 737
boron in steel, 737
Borax : see Sodium tetraborate
Boric acid : D. of; (ti) 3 1 1
Boron: as bis(salicylato)borate(III) anion, (se)

154

D. of, 154 .

by 1 , 1 '-dianthrimide, (s) 737
by ferroin, (se) 154

in steel, 737 ■(

Brass: analysis of, 803
Bromates, D. of: as silver bromide, (g) 491
iodometric, (li) 392
with standard arsenite soln., (ti) 391
(reeo/jo Potassium bromate solution) i '
Bromides, D. of: as silver bromide, (g) 491, (ti)
334 . .

by EDTA, (ti) 334
by mercury(I), (cm) 560.
by oxygen flask, 116
by silver ion, (cm) 562
by silver nitrate, (ti) 339
by Volhard’s method, (ti) 343
with iodide, (ti) 339
4-Bromomandelic acid: 426 .

Bromophenol blue: 240
Bromopyrogallol Red : 1 6 1 , 276, 322

903

INDEX

Bronsted bases: titration with strong acids, 257
Brensted theory of acids and bases: 15, 238
Buffer action: 42
Buffer capacity: 44
Buffer mixture: universal, 886
Buffer solutions: 42, 886 ,
acetic acid-sodium acetate, 45
for EDTA titrations, 322, 324, 325
preparation for NBS standards, 587
Bumping of solutions: 94
Buoyancy of air in weighing: 70
Burette: 78, 223
piston, 81
reader, 80
weight, 81

Burettes: calibration of, 80
manipulation of, 80
standardised, 79
stopcock lubricant, 79
tolerances of, 79

with automatic filling devices, 102, 103
with diaphragm plastic tap, 78
with glass stopcock, 78
with Teflon tap, 78
Burette holder: 79
Fisher, 79
Burners: 'Amur, 89
Bunsen, 83
electric Bunsen, 89
Meker, 89

Pittsburgh universal, 89
Tirril, 89

Cadmium, D. of: as metal, (eg) 536
as 2-naphthoquinoline complex, (g) 456
as quinaldate, (g) 457
by EDTA, (ti) 324
by fluorimetry, (fu) 778
by polarography, 653, 655
by pyridine method, (g) 457
by oxine, (ti) 393

Cadmium and zinc: separation of, on an anion
exchanger, 185

Cadmium liquid amalgam: 397
Cadmium sulphate, thermal analysis: 858
Calcein (fluorescein iminodiacetic acid):

chelating agent for calcium, 779
Calcichrome: 274, 329
Calcium, D. of: as carbonate via oxalate, (g)
437, 458

as oxalate, (g) 437, 458
as oxide via oxalate, (g) 437
as tungstate, (g) 458
by Calcein, (fu) 779

Calcium, D. of— continued
by EDTA, (p) 609, (ti) 320
by potassium permanganate, (ti) 354
by urea hydrolysis method, (ti) 354
in limestone or dolomite, (fl) 837
in presence of barium, (ti) 329
with CDTA, (ti) 329
with lead by EDTA, (ti) 325, (ass) 837
with magnesium by EDTA, 325
Calcium oxalate, thermal analysis: 858
Calcon: 273
Calculators: 127
Calibration: of apparatus, 124
of burettes, 80
of graduated flasks, 72
of pipettes, 87
of weights, 62

Calibration curve: in spectrophotometry, 728,
830

Calmagite: 274
Calomel electrode: 49
forms of, 569
potential of, 569, 890
preparation of, 570
Capacitative cell: 627
Carbon dioxide: adsorbents for, 492
Carbonate, D, of: (g) 492, (ti) 306
andwith hydroxide, (ti) 309
Carbonate and bydrogencarbonate, D, of: in
mixtures, (ti) 310
Carbosorb:492
Carrier gases: 209
Catechol Violet: 275
Cathode: 516, 639
Cation exchanger: 165
strongly acidic, 169, 171
weakly acidic, 169, 172
see also Ion Exchange

Caustic soda: analysis of commercial, (ti) 309
CDTA (trans-l,2-diaminocyclohexane-
A',/V,/V',/V',-tetra-acctic acid): 262, 328
Cells: barrier layer, 708
calculation of e.m.f. of voltaic, 49
capacitative, 627
concentration, 49, 567
conductivity, 619
H-type, 652

high frequency, 627, 628
in amperometric titrations, 675
in coulometry, 545
in electrolytic separations, 524
in Karl Fischer titrations, 687
in spectrophotometric titrations, 768
in potentiometric titrations, 567, 580, 591
in oxidation-reduction, 50

904

INDEX

Cells— ‘Continued . .

polarographic, 652

Cellulose for chromatography : 1 94, 206
modified, 195 .

preparation of, 203
Cellulose column : preparation of, 203
Centrifuge: 97 , .

use of, 97

Cerium, D. of: as cerium(IIl), (ti) 368 •
as oxide via iodate, (g) 458
by ammonium iron(II) sulphate, (ti) 369
by potassium permanganate, (ti) 368
by sodium thiosulphate, (ti) 377
Cerium(IV) ammonium nitrate: see Ammonium
cerium(IV) nitrate
Cerium(I V) ammonium sulphate : see
Ammonium cerium(IV) sulphate
Cerium(IV) hydroxide: 365
preparation of, 365

Cerium(IV) sulphate solutions: advantages of,
364

analyses involving, 366

formal potentials of, 364

indicators for, 365

oxidising properties of, 364

preparation of 0. 1 N, 365

standardisation of: by arsenic(III) oxide, 566

standardisation by iron, 368

standardisation by sodium oxalate, (ti) 366

Character of the deposit: (eg) 523
Chelate effect: 259
Chelates: 146, 216
for gas chromatography, 216
Chelating resins: 175, 190
Chelation: 145, 257
Chemical reagents: table of, 87 1
Chi-square distribution: 131
tables, 897

Chlorates, D. of: as silver chloride, (g) 493
by potassium dichromate, (ti) 363
by sodium thiosulphate, (ti) 381
Chloride and bromide, separation of: on an
anion exchanger, 182

Chlorides, D. of: as silver chloride, (g) 433, 494
(li)334

by mercury(I), (cm) 560
by mercury(ll) chloranilate, (s) 753
by mercury(Il) nitrate, (ti) 346
by mercury(ll) thiocyanate method, (s) 754
by oxygen flask, 1 16

by silver nitrate, (p) 604, (th) 868, (ti) 340
by Volhard’s method, (ti) 342
Chlorites, D. of: as silver chloride, (g) 494
Chloioplatinic acid : 477

Chromatographic columns: 193, 204, 206
plates, 196
development of, 197
Chromatography: adsorption, 10
apparatus for, 196
column, 193, 204, 206 ■
gas, 10, 209

high performance liquid, 198
introduction to, 10, 193
on cellulose, 202, 204, 206
on paper, 10, 193, 199, 200
partition, 10, 193
thin layer, 10, 196, 200
uses of, 1S2 etseq

Chromium, D. of: as barium chromate, (g) 459
as lead chromate, (g) 460
by ammonium iron(n) sulphate, (p) 606
by diphenylcarbazide, (s) 738 .
by iron(ll) via dichromate, (ti) 362
in chromite, (ti) 362
in chromium(III) salt, (ti) 361
in nickel alloy, (aa) 842
in steel, 739

Chromium and manganese: simultaneous D. of,
(s) 763

Chromium(in), D. of: with iron(lll), (ti) 331
CbromiumGlI) salt to chromate: 459
Chronopotentiometry: 633, 668
principles, 668
procedure, 669

Cleaning of glass apparatus : 7 3
Cleaning mixture: 73
Coagulation values: 406, 407
Cobalt, D. of: as mercurithiocyanate, (g) 461
by cyanide ion, (p) 61 1
by EDTA, (p) 609
by l-nitroso-2-naphthol, (g) 460
by nitroso-R-salt, (s) 739
in steel, (s) 740

Cobalt, sepn. of: on anion exchanger, 183
on cellulose column, 204
plus copper, nickel, zinc, (ch) 200
plus copper, nickel, (ch) 200
plus manganese, nickel, zinc, (ch) 199
plus- nickel, (ch) 204, (cm) 550, (p) 611
plus uranium, (ch) 188
Colloidal state: 404.

See also Lyophilic, Lyophobic .
Colorimeters:
double-cell type, 717
instruments, 707, 713
light filters for, 710
photoelectric, 713
single-cell type, 713
Colorimetric analysis : 6, 693

905

INDEX

Colorimetric analysis: —continued
criteria for, 727
general remarks on, 693, 727
theory of, 695
titration, 704

see also Spectrophotometry
Colour change interval : 237
Colour measurement: classification of methods,
699

Colours: approximate wavelengths of, 694
complementary, 711

Column chromatography ; iee Chromatography
Columns: in chromatography, 210
Combustion flames: 814
Common ion effect: 23
guantilative effects of, 27
Comparators: permanent colour standards, 703
Lovibond, 703, 704, 731
projection, 793, 794
Comparison of results; 130
Complementary colours : 7 1 1
Completeness of deposition: 321
Completeness of precipitation; 404
Complex ions: 30

D. of instability constants by polarography,
642

dissociation or instability constants of, 31
stablity constants of, 264
table of stability constants of, 263, 264
Complex formation reactions: 145, 224, 226
Complexes: chelation, 145
inert, 260

ion-association, 148
labile, 260
stability, 257

Complexomctric titratioos; general discussion,
224, 260,316, 601
see also EDTA
‘Complexones’: 261
Computers: 127

Concentrated acids: D. of strength, (ti) 308
Concentration cells: 49, 567
Condenser current: 636
Conditional stability constant: 264
Conduclimetry: 5
Conductivity: 61 5
limiting ionic at 25 °C, 6 1 6
measurement of, 616
molar, 616
specific, 615
table of, 616

variation during titration, 621
Conductivity bridge: 617, 620
instruments, 619
Conductivity cells: 619

Conductivity cells— continued
immersion type, 619
platinising of electrodes of, 6 1 9
Conductometric titrations: apparatus for, 617
applications of, 621
basis of, 617

experimental details for, 621, 625 '

general considerations of, 61 5
titration curves, 622
see also High-frequency titrations
Congo red: 240, 243
Conjugate acid (base): 16
Control determination : 1 24 >

Controlled cathode potential electrolysis: 524
Controlled potential electro-analysis: 538
apparatus for control of cathode potential,
626

auxiliary electrode for, 525
current and total applied voltage during, 525
D. of metals in alloys by, 539
evaluation of limiting potential for, 525
general considerations, 525
Controlled potential at the cathode; applications
of: D. of antimony, copper, lead and tin in
an alloy, 539

Copper, D. of: as benzoin-a-oximate, (g) 463
as copper(I) thiocyanate, (g) 462
as metal, (eg) 532

by biscyclohexanone oxalyldihydrazone, (s)
740

by ccrium(lV) sulphate, (t!) 367
by EDTA, (p) 639, (s) 740, 769, (ti) 321
by neo-cuproin, (se) 156
by potassium iodate, (ti) 389
by sodium diethyldithiocarbamate, (se) 155
by sodium thiosulphate, (p) 60S, (ti) 319
in brass, (em) 803
in copper(I) chloride, (ti) 367
in copper ore, (ti) 380
in crystallised copper sulphate, (ti) 367, 379
in steel, (s) 74 1 , (v) 683 ,
in white metal, (em) 807
Copper, sepn. of: from brine by ion exchange,
190

plus cobalt, (ch) 201
plus cobalt, nickel, zinc, (ch) 200
plus lead, (v) 656
plus nickel, (ch) 206, (eg) 537
Copper-nickel alloy: analysis of, 538
Copper sulphate, thermal analysis: 858, 862
Co-precipitation: 409
Costs: of equipment, 7, 9
Coulomb: 516, 542
Coulometer: hydrogen-nitrogen, 543
hydrogen-oxygen, 543

906

INDEX

Coulometer— con/mi/et/ . ‘ :

iodine, 543 - ’ -

silver, 543

Couloraetric analysis: general discussion, 5, 542
Coulomefric titrations: 542, 548
advantages of, 549
circuit and cell for, 552
commercial titrators, 556
constant current sources, 551
current measuring devices, 552
detection of end points in, 550
electrolytically generated reagents for, 553
experimental details for, 556
external generation of titrant, 555
fundamental requirements for, 549
general discussion of 548
instrumentation, 551
integrating motor-counter, 552
low inertia integrating motor, 556
primary, 542
secondary, 542
time measurement, 552
Coulometry at constant current: see
Coulometric titrations
Coulometry at controlled potential : 542, 543
current in, 544
electrolytic cells for, 545
general technique of 547
separation of cobalt and nickel by, 548
Counter e.m.f : 517
Counter ions: 406
Crucibles: cleaning of 1 1 1
cooling of 92, 114
Gooch, 109
heating of 1 14
perforated screens for, 115
porcelain, 84

porcelain filtering, 96, 110
preparation of 110
Rose, 461

sintered filtering, 96, 1 10
tongs, 96

Vitreosil filtering, 96
with permanent porous plates, 96, 1 10
Crushing and grinding: 137
Cupferron: 150,421,466,484,487

Cupron: 424, 463

Copperfl) chloride: analysis of (ti) 367

Current: density, 516
diffusion, 637
efficiency, 516

limiting (with dropping mercury electrode),
636

migration, 636
residual, 636

Cnrrent'Voltage curve: 634 ' ’ ■

Cyanide: D. of cobalt by, (p) 61 1
D, of nickel by, (p) 61 1 '

titration with silver nitrate, 260, 345 ' .
Cyanide, D. of: as silver cyanide, (g) 494
with silver nitrate, (ti) 345
Cyclohexane-1, 2-dione dioxime: see Nioxime

Damping of balances :'62 . .

Daniell cell: 50
d.c. arc source : 79 1 , 792, 802
Dead-stop end points: 685
Decantation: 112
Decomposition potential: 517
Degreasing agent : 74
Delves cup: 819
Demasking agents: 267, 330
Depolariser: anodic, 527
cathodic, 520

Deposit : character of (eg) 523 •

Deposition : completeness of 52 1
potentials of metals, 523
with controlled potential at the cathode, 547
Desiccants: 91
Desiccators: 91
cooling of crucible in, 92
desiccants for, 91
Scheibler type, 92
uses of 92
vacuum, 92
Detection limit (fl): 834

Detectors: for gas chromatography, '21 1 , .

for spectrometry, 700, 707, 708, 820 .
Determinations: blank, 124
control, 124 , ■ -

parallel, 125, 132 ■

Devarda’s alloy: 314, 730
Deviation: mean (average), 127
standard, 128

1.2- Diaminocyclohexane-A',iV,A'',/V'-tetra acetic

acid (DCTA or CDTA): 262, 328
l,l'-Dianthrimide: 737
Dichlorofluorescein: 337
Dichromate ion: D, of coulometrically, 559
Diethyl oxalate: in homogeneous pptn., 412

1.2- DianunoethanetetTa acetic acid: see EDTA
Differential scanning calorimetry: 849, 859

application, 862
experimental, 862

experimental and instrumental factors, 861
instrumentation, 860
Differential thermal analysis: 849, 859 -
application, 862
experimental, 867

experimental and instrumental factors, 865 ■

907

INDEX

DifTerenb'al tJiennal analysis— conf/wuerf
instrumentation, 860
Differential titration: 593
Diffraction gratings: in spectrophotometers,

712

theory of, 7i2

Digestion: of precipitates, 410
vessels, 106

Di-iododimethyifluorescein: 339
Dilution method: 699

Dimethylglyoxime: 150, 161,420,447,474, 748
sodium salt, 421

Dimethyl oxalate: in homogeneous
precipitation, 412

2,9-Dimclbyl-l,10-phenanlbroline: jce Neo-
cuproin

2,3-Dimethyl-l-phenyI-5-pyrazolone : see
Phenazone

Dimethyl sulphate: in homogeneous
precipitation, 413
Dipbenylamine: 294

Diphenylamine-p-sulphonic acid, sodium salt:

359, 360, 363
Diphenylbenzidine; 294

Diphenylcarbazide: as adsorption indicator, 345
as colorimetric reagent, 738
DiphenyllUacarbazonetsee Dithizone
Direct reading emission spectrometer: 800
instruments, 800
operation, 800
Dispensers (liquid): 103
Displacement titrations: 253
borate ion with a strong acid, 253
carbonate ion with a strong acid, 254
Dissociation (ionisation) constant: 22
calculations involving, 24
D. of, 19

D. of for a complex ion, 642
D. of for an indicator, (s) 761
of polyprotic acids, 22
tables of values in water at 25 “’C, 888
true or thermodynamic, 18
Distribution coefficient: 143, 173
and per cent extraction, 143, 173
Dithiol: 151, 747, 749, 751
Dithizone: 150, 158

DME: see Dropping mercury electrode
Draining time: 77
Drop time: in polarography, 650
Dropping mercury electrode: 650, 675
Dry box: 93
Drying reagents: 9 1
comparative efficiencies, 91
Drying of precipitates: 91
Drying pistol; 93

Duboscq colorimeter: 707
Duplication method : 70 1 , 706

EDTA: 261

alkalimetric titrations, 267
back titrations, 266
buffer solutions for titrations, 324
direct titrations, 266
high-frequency titrations with, 629
ionisation of, 261
masking agents for titrations, 267
metal indicators for titrations, 269, 271
miscellaneous titrations, 267
potentiometric titrations with mercury
electrode, 607

preparation of magnesium complex, 320
preparation of standard solutions of, 3 17
purification of, 317

replacement or substitution titrations, 266
stability constants of metal complexes, 263,
264

standardisation of solutions of, 317
structural formula of metal complex, 263
summarised procedures, 324
thermal analysis of, 859
titration curves, 265
titration of mixtures with, 267
types of titrations, 266
EGTA:262, 327
Electric: hot plates, 90
ovens, 90
stirrers, 90
Electrical units: 515
Electrification due to wiping: 69
Electro-analysis: see Electrolysis and
Electrogravimelry
Electrochemical: equivalent, 516
series, 49

Electro-deposition: completeness of, 521
Electrode potentials: 46, 880
change of, during titration, 288 .

Nernst equation of, 47
reversible, 49
standard, 48
table of standard, 48
Electrode reactions; 518
Electrode vessel: 530
Electrodeless discharge lamps : 8 1 9
Electrodes: antimony, 572
auxiliary, 525, 545, 563
bimetallic, 592
calomel, 49, 526, 569, 880
dropping mercury, 650, 675
generator, 553
glass, 577, 579

908

INDEX

Electrodes— con/i/iMerf
graphite indicating, 681
hanging mercury drop, 664, 665
Hildebrand, 568
hydrogen, 568, 572
indicator, 563, 566, 57I;*593
ion sensitive, 575
isolated generating, 563
Lindsey, 568
liquid membrane, 577
membrane, 577
mercury (cathode), 529
mercury/mercury(II)-EDTA, 607
non-polarisable reference, 635
of the firsrkind, 566
of the second kind, 566
platinising of, 619
platinum, 527
polarisable micro-, 635
polarised, 603

potentials of common reference, 890
reference, 568, 890
rotating micro-, 682
silver-silver chloride, 49, 571, 890
spectrographic, 802, 805
working, 525, 545, 634
see also Ion sensitive electrodes
Electrodes for electrolytic determinations:
Fischer, 528
Mercury cathode, 529
platinum gauze, 527
rotating anode, 528
use and care of, 531
Electrogravimetry: 5, 515
Electrolysis: apparatus for, 527
commercial forms of, 528
constant current, 527
controlled potential, 525, 538
internal, 540
laws of, 515
rapid, 534

separation of metals by, 522
slow, 533
technique of, 527
Electrolytes: strong, 15,20
weak, 15

Electrolytic ceUs: 516
Electrolytic dissociation : 14
Electrolytically generated reagents: bromine,
558

cerium(lV) ion, 550,560
tron(H) ion, 559
cyanoferrate(II) ion,
hydrogen ion, 562, 563
hydroxide ion, 562, 563

Electrolytically generated reagents— continued
iodine, 557, 558 . ' •
mercury(I) ion, 560

silver ion, 562 .■ . ■ ,

Electrolytic separation of metals: 522
of cobalt and nickel with controlled cathode
potential, 548

of copper and nickel, 537 ' ■

see also under individual metals ■ ■
Electromagnetic radiation: 694
Electron as standard reagent: 549
Electron capture detector: 2 1 3
Electrothermal atomisers: 817
Elution: 173, 193, 198,210
E.m.f. : of voltaic cells, 49, 579 ’
polarisation, 517

Emission spectrographic analysis: advantages
of, 789, 797

apparent disadvantages, 789
applications of, 788
comparison sample method, 799
D. of copper and lead in white metal by, 807
D. of lead in brass by, 803
direct reading instruments, 800
electrodes for, 792, 802
equipment for, 789, 805
excitation sources for, 791
general discussion of, 6, 787
internal standard method, 799
of a complex inorganic mixture, 801
of a non-ferrous alloy, 801
photographic details for, 798, 802
qualitative, 795, 801
quantitative, 798
theory of, 787, 810
End points: 223
dead-stop, 685

in amperometric titrations, 676, 685
in coulometric titrations, 550
in EDTA titrations, 318
in neutralisation reactions, 236
in oxidation-reduction reactions, 292, 372,
387

in precipitation reactions, 281
indicators for, 236

location by analytical (derivative) methods in
potentiometric titrations, 596
Eosin: 284, 776
Equilibrium constants: 18
of EDTA reactions, 264
of redox reactions, 53
Equivalence point: see End points
Equivalent weights: 225, 226
of acids, 226
of bases, 226

909 -

INDEX

Equivalent weights— confinuerf-
of oxidising and reducing agents, 230, 231
of salts, 226

partial ionic equations for calculation of, 230
variation with reaction, 233
see also Normal solutions
Eriochrome Black T: 272, 274, 745
Eriochrome Blue-Black B: 273
Eriochrome Blue-Black KC; 273
Eriochrome Cyanine R: 729
Errors: absolute, 127

additive and proportional, 123
classification of, 123
determinate or constant, 1 23
in weighing, 69

indeterminate or accidental, 123
instrumental and reagent, 123
mim’misation of, 124
of method, 123
operational and personal, 123
relative, 127

Ether (diethyl): drying precipitates by, 457, 471,
473,478,480,489, 507
extraction ofiron as iron(ni) chloride, 157
p-Ethoxychrysoidine: 391, 392
Ethylenediamine : 428

Ethylenediaminetctraacctic acid (disodium salt):
see EDTA

2, 2'-Ethylenedloxybis{ethyUminodi (acetic
acid)} -.see EGTA
Evolution methods: 430
External indicators: 295
Extinction: 696
Extinction coefficient: 696
determination of, 697
molar, 697
specific, 697

Extraction: see Solvent extraction

F-test: 131
tables, 896

Faraday constant: 46, 542
Faraday’s laws: 515, 542
Fast Sulphon Black F: 275
Ferric alum indicator: see Ammonium iron(Ill)
sulphate

Ferric iron: see Iron(HI)

Ferricyanides: see Hexacyanoferratc(III)
Ferrocyanides: see Hexacyanofcrrale(II)
Ferroin: 154, 294, 366
modification by substituents, 294
preparation of indicator solution, 154, 293
Ferromanganese: analysis of, (ti) 329
Ferrous ammonium sulphate: see Ammonium
iron(ll) sulphate

Ferrous iron: see Iron(II)

Filter papers: 107
folding of, 108 •
incineration of. 1 13
macerated, 109
quantitative, table of, 107
Filter pulp: 109
Filtering crucibles: 95
Filters, optical: 693, 710
Filtration: 95, 107
accelerated, 109

technique of, 108, 112 • ,

with filter papers, 95
with filtering crucibles, 109, 110
Flame emission spectroscopy : 8 1 0
Flame ionisation detector: 212
Flame photometer: 8 10, 824
instrumentation, 814, 824
Flame photometric analysis; data for common
elements, 834
D. of

calcium, lithium, potassium and sodium,
837

calcium and magnesium in tap water, 837
chromium in nickel alloy, 842
trace lead in ferrous alloy, 840
vanadium in lubricating oil, 840 .
sulphate ion (Atomic Absorption
Inhibition Titrimetry), 843
elementary theory of: 81 1
evaluation methods, 830
experiments in, 835
general discussion, 810
instrumentation, 814, 824
non-flame techniques, 817
safety practices, 827
vapour technique, 819
Flame photometry: 6, 285
Flame spectrometry : 8 10
Flames: temperatures of, 815
Flasks: conical (Erlenmeyer), 84
graduated, 74
Flocculation of colloids: 406
Flocculation values: 406
Fluorescein: 284, 337, 776
Fluorescence; 773
and concentration, 773
Fluorescent indicators: 776
Fluorides, D. of: as lead chlorofluoride, (ti) 343
as triphenyl tin fluoride, (g) 494
by ion sensitive electrode, (p) 589
by lanthanum acetate, (hf) 630
by null point method, (p) 612
by thorium nitrate, (am) 689
with aid of cation exchanger, 186

910

INDEX

fluorides, D. of— conrmuerf . -

with thorium chloranilate, (s) 753 .
Fluorimeters:774 •

Fluorimetric analysis: applications of, 776
general discussion of, 6, 773
determinations by, 777
instruments for, 774
Fluorosilicates, D. of: (g) 495
Fluxes: 105

Formal potentials : 29 1 , 364 '

Formation constant: 31
Fractional predpitation: 29
Fuchsine (rosaniline) : 39 1
Fuel gases for flame photometry : 8 1 5
Funnels: 84
analytical, 96
Buchner, 96
Hartley, 96
sintered glass, 96
slit sieve, 96
Furii-<x-dioMme: 421
Fusions: with lithium metaborate, 105
with sodium carbonate, 502
with sodium hydroxide, 503

Gamma, of a photographic emulsion : 798
Gas chromatography: 209
detectors for, 21 1
ofmetal chelates, 216
quantitative analysis by, 214
using programmed temperature, 215
Gaussian distribution: 129
Gelatin: 653
Gels: 405

Glass elKtrode: 573, 575

forms'of, 576
glasses for, 576
use of, 573
Glassware: 84
cleaning of, 73
graduated, 71
Glove box: 93
Glycerol; 311, 441
Gold, D. of: as metal, (g) 464

Gooch ctueiUes; 95, 11^
asbestos for, 95
funnels for, 109

heating of, 110
preparation of, 1 10
rubber sleeves for, 1 10
Vitreosil (silica), 85, 1 13
Graduated glassware: 71
burettes, 78

calibration of, 74, 75, 80 ' ' ■

Graduated glassware— con/mueri
cylinders, 82
flasks, 74
pipettes, 76

temperature standard for, 72
units of volume, 71
Gram: 61

Graphite rod and Graphite tube : for atomic
absorption spectroscopy, 817, 818
Gratings :'jee Diffraction gratings
Gravimetric analysis : 5, 403
calculations of, 431
general discussion of, 403, 43 1
physical forms for precipitation and
weighing, 403
simple, 432
systematic, 449
technique of, 103, 112
theory of, 403, 43 1
Grinding: 137

Hafnium: extraction by liquid ion exchanger,
178

Half-wave potentials: 638
D. of the cadmium ion, 653
table of, 891 ' •

use of in D. of formula and stability constant
of a metal ion complex, 642
HaUdes in admixture, D. of: by absorption
indicators, (ti) 339
by an indirect method, (ti) 340
with the aid of an ion exchanger, 182
Hanging mercury drop electrode: 664, 665
Hardness of water, D. of: by EDTA, (ti) 328
permanent, 328
temporary, 328
total, 328

Hartmann diaphragm: 790, 793
Heating apparatus: 88
Hehner cylinders: 705
Hexacyanoferrate(II): D. of, (ti) 370
Hexacyanoferrate(in): D. of, (ti) 385
Hexamethylenetetramine (hexamine): 324, 326,
333

High alumina cement, DTA and DSC studies:
863

High-frequency titrations: cells for, 627
construction of apparatus for, 628
determinations by, 629
general considerations of, 626 ■

High performance liquid chromatography : 1 98
see also Chromatography
High voltage a.c. spark : 793
Hollow cathode lamp: 819 ■ •

911

INDEX

Homogeneous precipitation : see Precipitation
Hotplates: 90, 95

Hydrazine: as anodic depolariser, 546
D. of ; by potassium iodate, (ti) 389
Hydrochloric acid: composition of constant
boiiing, 297

HCl content of concentrated acid, (ti)

322

pH of 0.1 A/,887

preparation of constant boiling, 297
preparation of 0.1 A/, 297
preparation of O.OIA/, 298
standardisation of

by a high frequency titration, (hf ) 630
by an idiometric method, (ti),301
by sodium carbonate, (ti) 298
by sodium tetraborate, (ti) 300
gravimetrically as silver chloride, 435
Hydrogen electrode: 568, 572
Hildebrand bell-type, 568
Lindsey, 568

preparation of typical, 568
standard or normal, 47
use of, 568

Hydrogen ion concentration: 20, 33, 245
in buffer solutions, 42
in hydrolysed salts, 38
of some common reagents, 887
see also under pH
Hydrogen ion exponent: 34
see also under pH
Hydrogen overpotential: 519, 520
Hydrogen peroxide, D. of: by cerium(IV)
sulphate, (ti) 369

by potassium permanganate, (ti) 355
by sodium thiosulphate, (ti) 381
Hydrogen sulphide, D. of: (ti) 384
reductions with, 401
Hydrolysis: 37
of salts, 38

Hydrolysis constant: 38
and degree of hydrolysis, 38
Hydroxide, D. of: 302
D. of with carbonate, 309
Hydroxides: precipitation at controlled pH,
419

Hydroxom'umion: 14

Hydroxylamine, D. of: by potassium bromate,
(ti) 394

Hydroxyl ion exponent, pOH: 35
4'Hydroxyphenylarsonic acid : 427
8'Hydroxyquinaldine: sec 2-Methyloxine
8'Hydroxyquinoline:see Oxine
Hypochlorites, D. of: (ti) 383
as silver chloride, (g) 494

Hypopbospbites, D. of : as ammonium

magnesium phosphate or as magnesium
pyrophosphate, (g) 496
by mercury(II) chloride as mercury(I)
chloride, (g) 496

Ignition: of crucibles, 112, 114
of hydrated salts, 432
of precipitates, ill, 432
Ilkovic equation ; 637
Immersion beater: 89
Indicator electrodes: 566
dropping mercury, 650, 675
rotating micro-, 682
Indicator constant: apparent, 237, 239
Indicators: acid-base, 236, 240
adsorption, 284

choice in neutralisation reactions, 255

colour change interval, 240

colours and pH ranges, 240

D.ofpKof,76I

external, 295

fluorescent, 284

in redox titrations, 292

internal, 365

metal ion, 269

mixed, 243, 299

neutralisation, 236

prep, of solutions of, 242

reagent serving as own, 295

screened, 243

starch, 373

table of acid-base, 240
table of adsorption, 287
table of redox, 294
theory of action of acid-base, 236, 237
universal or multi-range, 244
Inductive cell: 628
Infrared: lamps and heaters, 90
radiation, 694

Instability constants: D. of, (v) 642
Integration: automatic, 215
by weighing, 215
Interference Alters : 7 1 0
Interferences: in analysis, 8
in flame spectrometry, 821
Internal electrolysis: 540
Internal indicators: 365
Internal standards: 365
lodates, D. of: as silver iodide, (g) 496
with standard arsenite solution, (ti) 387
Iodides, D. of: as palladium(n) iodide, (g) 495
as silver iodide, (g) 496
by EDTA, (ti) 334

912

INDEX

Iodides, D. of— con/muerf

by mercury(ll) nitrate, (am) 680

by mercury(I) ion, (cm) 560
by silver ion, (cm) 562
by siiver nitrate, (p) 605, (ti) 34 1
by Volhard’s method, (ti) 340, 342
with bromide, (ti) 339
with chloride, (p) 605, (ti) 339
lodimetry and iodometry : 370
detection of end point in, 372
general discussion of, 370
sources of error in, 372
Iodine : purification by sublimation, 99
Iodine solution: indicators for, 372, 373
oxidising properties of, 370, 37 1
prepn of 0.1 IV, 379

standardisation by arsenicflll) oxide, 378
standardisation by sodium thiosulphate, 379
lodometric method for standardising strong
acids: 372

Ion activity meters: 584
Ion-association complexes: 148
Ion-electron method: 228
rules for balancing ionic equations, 229
Ion exchange: 9, 165, 195, 198
capacity, 170, 180
chromatography, 172
columns, 178
liquid, 177

methods, 171, 172, 175
preparation of, 165, 166
equilibria, 169, 173
experimental techniques, 178
Ion exchange resins: 165
action of, 167
anion, 165
capacity of, 170
cation, 167
chelating, 175
D. of capacity, 170, 180
distribution coefficient, 173
separation factor, 173
table of commercial, 168
uses of, 175, 198, 302, 758
volume distribution coefficient, 174
Ion sensitive electrodes : 57 1 , 575
commercially available, 578
design of, 576
fluoride, 589

glass composition for, 576
meters for, 584
selectivity, 577
Ionic equations: 229, 230
for calculations of equivalent weights, 225
ot oxidants and reductants, 227

Ionic molar conductivities (Table) : 6 1 6
Ionic product of water: 33
table at various temperatures, 34
Ionic strength: 19, 20
Ionic strength adjustment buffer: 589
Ionisation constants: of indicators, 237
table of, 888

see also Dissociation constants
Ionisation suppressant: 823
Iron(Il), D. of: by cerium(IV) ion, (cm) 560
by cerium(IV) sulphate, (ti) 370
by potassium dichromate, (ti) 360
by potassium permanganate, (ti) 354
see also under Iron

Iron(III), D. of: by EDTA, (p) 610, (s) 710, (ti)
322

with chromium(III), (ti) 33 1
see also under Iron
Iron(III) indicator solution: 341
Iron(III) rednction of: by hydrogen sulphide,
401

by Jones reductor, 395
by liquid amalgams, 397
by silver reductor, 398
by tin(II) chloride, 399
by sulphurous acid, 400
Iron, D. of: as iron(III) oxide, (g) 440, 465, 466
as oxinate, (se) 158
by basic formate method, (g) 465
by cerium(IV) sulphate, (ti) 366, 370
by chloride extraction, (se) 157
by cupferron, (g) 466
by 1,10-phenanthroline, (s) 742
by potassium dichromate, (ti) 360
by potassium permanganate, (ti) 354
by thiocyanate, (s) 741
by thioglycollic add, (s) 743
in amponium iron{III) sulphate, (g) 442
in brass, (em) 803
in iron ore, (se) 157, (ti) 361
in presence of hydrazine hydrate, 443
in steel, (se) 157, (ti) 361
Iron and aluminium: separation on a cellulose
column, (ch) 202
D. with EDTA, (s) 771
Iron ore: analysis of, (ti) 361
Iron: ware, 88
wire, 236

Isotopic dilution: 125

Jones reductor: 395
applications of, 396
limitations of, 396
preparation of, 395
uses of, 397

913

INDEX

Journals of analytical chemistry: 5, 13

Karl Fischer reagent; 687
apparatus and experimental detail, 687
D. of water content of sodium acetate, 689
interfering substances in, 688
preparation of, 687
standardisation of, 689
theory of titration with, 687
Kilogram: international prototype, 61
Kinetic masking : 269, 33 1
Kinetic methods: 7

Kjeldahl’s method for nitrogen detn. : 3 1 2

Lambert’s Law: 695
Lanthanoids: 172, 483

Lanthanum ion: titration with EDTA, (hf) 630
Lead, D. of: and calcium by EDTA, (ti) 328
and copper, (v) 656
and tin by EDTA, (d) 333
as chromate, (g) 444, (ti) 362
as dioxide, (eg) 535
as molybdate, (g) 467
as salicylaldoximate, (g) 468
as sulphide, (s) 744
as dithizone, (sc) 158, (s) 744
by EDTA, (p)610
by potassium dichromate, (am) 676
in bearing metal, (eg) 539
in brass, (em) 803, 804
in ferrous alloy, (aa) 840
in solder, (ti) 333
in steel, (v) 656
in tartaric acid, (s) 744
in white metal, (em) 807
Lead liqm'd amalgam: 397
Lewis acids and bases: 17, 258
Liebig’s method for cyanide: 260
Deniges modification, 261
Ligand: 146, 257

Light filters: for colorimeters, see Filters,
optical

Limiting cathode potential: 525
see also Controlled potential electro-analysis
Liquid amalgams: applications of, 397
apparatus for reductions, 398
general discussion, 397
reductions with, 398
zinc amalgam, 397
Liquid ion exchangers: structure. 177
uses, 189, 577

Liquid junction potential: 567
Literature of analytical chemistry: 12, 13

Lithium, D. of: admixture with sodium and
potassium, (g) 468
as aluminate, (g) 469
Litmus: 241
Litre :xxvii, 71

Littrow mounting: 71 1, 721, 723
Logarithms : four figure, 898
Long cell absorptiometer: 7 1 5
Loribond comparator : 704
Nessleriser, 703
Low voltage d.c. arc: 792 ■

Lubricants for glass stopcocks : 79
Lyophilic colloids: 405
Lyophobic colloids: 406
stability of, 406

Macerated filter paper: 109
Magnesium, D. of: as ammonium magnesium
phosphate, (g) 444
as pyrophosphate, (g) 444
by EDTA, (p) 609, (ti) 325, 327, 329
by Eriochrome Black T, (s) 745
by 2-methyloxine, (g) 469
by Titan Yellow, (s) 745
with calcium, (aa) 837, (ti) 325, 327
with manganese and zinc, (ti) 330
Magnetic stirrer: 94
Mandelic acid; 426

Manganese, D. of: as ammonium manganese
phosphate or as pyrophosphate, (g) 470
by EDTA, (ti) 330, 331
by periodate oxdn., (s) 746
by permanganate, (p) 606, (ti) 336 .
in brass, (em) 803
in ferromanganese, (ti) 33 1
in pyrolusite, (p) 607
in steel, (p) 607, (ti) 358
with iron, (p) 607, (ti) 358
with magnesium and zinc, (ti) 330
Mannitol: 312, 603
Masking agents: 8, 267
Mass action law : 1 7
application of electrolyte solutions, 18
Matrix effects: 823
Maxima in polarograpby : 638
suppression of, 638, 653
Mean deviation: 127
relative, 127
standard, 127
Measuring: cylinders, 82
flasks. 74

see also Graduated glassware
Membrane electrode : 57 1 , 577
Meniscusrteading position of, 80

914

INDEX

Metcaploacetic acid : jee Thioglycollic add

Mercury:purificationof, 650-
Mercury, D. of: as sulphide, (g) 470 ,

3s‘ihionalide,(g)471 ■

byEDTA,(p)610,(ti)325
by oxygen flask, 116
by potassium iodate, (ti) 388 ,

Mercury cathode: 529
cells, 529

deposition of metals at controlled potential
on, 530 : ■

Mercury(II) chloranilate: 754 •

MercuryPl) nitrate : standard soln.. of, 346
Metcury/mercury{II)-EDTA electrode (mercury
electrode); 607 , , . .

potentiometric titration of metallic ions with
EDTAand,607
prepn. of, 608
Mercury thiocyanate: 754
Metaphosphoric acid: in homogeneous
precipitation, 412
Metal apparatus: 85

Metal ion indicators in EDTA titrations : 269
detection of colour change by instrumental
methods, 318
examples of, 271
general properties of, 269
requisites of, 269
theoryofvisualuse,270
Methylcellulose: 638
4-Methylnioxime : 42 1
^Methyl-oxine: 429, 469, 479
Methyl Orange: 240, 242, 391
-indigo carmine, 298
-Xylene Cyanol FF, 243
Methyl Red : 240, 242, 299, 39 1 , 76 1
Methyl thymol Blue: 277

Methyl Yellow: 240, 243
Microelectrodes: 635 . . ,

Micrometre: 694
Microphotometer: 794
Microwave oven: 90 . ,

Migration current: 632
Mixed indicators: 243
Mobile phase: 10, 193, 209

MobiUties, ionic; table oflimiting, at25°C, 616
Modulation: 821

Mohr procedure: experimental details of, 336,
338

Molar absorption coeflflcient : 697
Molar conductivity: 616, 617

Molar extinction coefficient: 697
Molar solution: definition, 226.

Molarity: 225
Mole: 225

Molybdate, D. of: (ti) 367
Molybdenum, D. of: as lead molybdate, (g) 471
as molybdyl oxinate, (g) 472
by dithiol, (s) 747

by silver reductor and cerium(IV) sulphate, ■
(ti) 367

by thiocyanate method, (se) 160
in steel, 160

Monochromator : 7 1 1 , 7 1 2, 820
Mortar: agate, 137
mullite, 137
percussion, 137
synthetic sapphire, 137
Muffle furnace: 90
Multiple range indicators; 242
Murexide:271

Naphthalene Black 12B: 391
1-Naphtholphthalein: 240, 242
N-Benzoyl-TV-phenyl hydroxylamine: 422
N.B.S. standards for pH: 587
Nebuliser-burner system : 8 1 4, 8 1 5
Neo-cupferron: 422
Neo-cuproin: 156

Nephelometer: instrumentation, 782
uses of, 78 1

Nepheloraetry: D. by, 784
general discussion on, 78 1
Nernst equation: 47, 525, 566, 576
Nessler tubes : 70 1 , 704
stand for, 701
Nessleriser: 703

Nessler’s reagent: prepn. of, 73 1
use of, 731

Neutralisation curves: 244
of a polyprotic acid with a strong base, 25 1 ,
256

of a strong acid and a strong base, 244, 249
of a weak acid and a strong base, 247, 255
of a weak acid with a weak base, 251, 256
of a weak base and a strong acid, 250, 256
Neutralisation reactions: 224, 600, 602
choice of indicators, 236, 255
Neutral red: 240, 243
Nickel, D. of : as metal, (eg) 538
by cyanide ion, (p) 61 1
by dimethylglyoxime, (am) 678, (s) 747, (se)
161, (g) 447

by EDTA, (hf) 630, (p) 609, (s) 771, (ti) 322
by oxine, (ti) 393

by pyridine-ammonium thiocyanate method,
(g)473

by salicylaldehyde oxime, (g) 473
in brass, (em) 803.

915

INDEX

Nickel, D. oI~conlinued
in nickel steel, (g) 448, (s) 748, (ti) 332
in presence of copper, (g) 473, 538
in steel, (s) 748

Nickel, sepn. of: plus cobalt, (ch) 204, (cm) 548,
(p)6U

plus cobalt, copper, (ch) 20 1
plus cobalt, copper, zinc, (ch) 200
plus cobalt, manganese, zinc, (ch) 199
plus copper, (ch) 206, (eg) 537
Nickel ware: 88
Nioxime:421,474
4'methyl-, 421

NITA (nitriloacetic acid): 262
Nitrates, D. of: as nitron nitrate, (g) 497
by ammonium iron(n) sulphate, (am) 686
by reduction to ammonia, (ti) 314
Nitriloacetic acid : 262

Nitrites, D. of: by cerium(IV) sulphate, (ti) 369
by diazo method, (s) 755
by potassium permanganate, (ti) 356
Nitrogen, D. of: by Kjeldahl’s method, 312
d-Nitrobenzene-azo-orcinol : 735
Nitron: 426, 491, 497
4-NitTophenol: 240, 243
l>Nitroso-2>naphthol: 154, 425, 460
2>Nitroso-I'naphthol: 740
N-nitroso-N-phenylhydroxylamine: see
Cupferron
Nitroso-R-salt: 739
Nitrous acid: removal of, 532
Normal distribution; 129
Normal solutions: 225
Normality: 102, 300
Notebook; laboratory, 58
Null-point potentiometry: 567, 612

Observations, recording of: 58, 307
Ohm: 515
law of, 515, 615
Oleum: 307
Opacity: 696

Optical density : see Absorbance
Optical filters: 693, 710
Organic chemical reagents: 87 1
Organic precipitants: 419
Osmium tetroxide catalyst; 366
Ostwald’s dilution law: 20
Overpotential: 5 19
hydrogen, on various cathodes, 520
Overvoltage: 519

Oxalates, D. of: as calcium carbonate via
oxalate, (g) 498

as calcium oxalate monohydtate, (g) 498

Oxalates, D. of— continued
as calcium oxide via oxalate, (g) 498
by cerium(IV) sulphate, (ti) 370
by potassium permanganate, (ti) 352
Oxidant; equivalent of, 230 •

Oxidation: 228
Oxidation number : 23 1
rules for determination of, 231
Oxidation number method: 231
Oxidation-reduction cells: 50
Oxidation-reduction curve: 289, 290
Oxidation-reduction indicators: table of, 294
Oxidation-reduction reactions: 53, 225, 227,

. 288,348,601,639
change of potential during, 288
equilibrium constants of, 53 . , -
indicators for the detection of end points in,
292

Oxidising agents: 228
equivalent weights for, 230 . .

table of, 232

Oxine: 150, 158, 163,422,472,487, 779
conditions for use, 423
D. of, with bromine, (cm) 558
for D. of metals
by potassium iodate, (ti) 393
by bromine, (cm) 558
preparation of reagent, 423
Oxygen: dissolved, effect in polarography, 643,
655

Oxygen flask; use of for elemental analysis, 1 15

Palladium, D. of: as dimethylglyoximate, (g)
420, 474

as nioximate, (g) 474
by EDTA, (ti) 325
Paper chromatography; 193
see also Chromatography
Paper strips: separations on, 201
Parallax: errors due to, 80
Parallel determinations; 123, 130
Partial ionic equations: 229, 230
Partition chromatography: 10, 193
Partition coefficient: 143
Patton and Reeder’s indicator: 273, 321, 326
Peptisation; 406, 407

Perchlorate, D. of: as silver chloride, (g) 488
Percussion mortar: 137
Periodates, D. of: as silver iodide, (g) 496
Peroxides: analysis of, (ti) 355
Persistent lines: 788
table of, 892

Persulphates, D. of: by cerium(IV) sulphate, (ti)
369

potassium permanganate, (ti) 357

916

INDEX

pH: 33 ■

at equivalence point in titrations, 245, 247,
250,251

BritisH standard of, 586, 886
calculations involving, 35
definition of, 35 ■

D. by colorimetric methods, 702
D. of by potentioraetric methods, 585
measurement of, 588.

of acetic acid-sodium acetate mixtures, 45,
886 ' • :
of common reagent solutions, 887
ofhydrolysed salts,'39, 41, 42
of hydroxide precipitations, 417
of NBS standards, 587, 594
secondary standards of, 886
pH meters: 582
direct reading type, 583 •
instrumentation, 583
operation of, 584
pHj.- 147
pK:36,238

D. of an indicator, (s) 76 1
tables of, for acids, 888
tables of, for bases, 889
pK,:36

pM indicators: 269
pOH:32

1,10-Phenanthroline: 161, 293, 294, 742
l|IO-Phenanthroline iTon(II) sulphate; see
Ferroin

Phenazone:485, 486
Phenolphthalein: 240, 241, 242
Pbenosafranine: 287

A’-Phenylanthranilic acid; 359, 360, 362
prepn. of indicator solution, 361
Phenylarsonic acid: 427

Phosphates, D. of : as ammonium magnesium
phosphate or as magnesium
pyrophosphate, (g) 498
as ammonium molybdophosphate, (g) 499
as quinoline molybdophosphate, (ti) 314
by EDTA, (ti) 335

by molybdenum blue method, (s) 756
by Phosphovanadomolybdate method, (s)

756 ■

nephelometrically, 785

PhospUtes, D. of: as ammonium magnesium
phosphate or as magnesium
pyrophosphate, (g) 500 •
as raetcuryd) chloride by mercurydl)
chloride, (g) 500
i^nosphorescence: 773

Phosphoric acid: action on indicators, 307
■ m the commercial acid, (ti) 307

Phosphoric adi— continued
neutralisation of, 307
Phosphorous acid: as reducing agent, 388
Phosphorus, D. of: by oxygen flask, 1 16
in brass, (em) 803
see also Phosphates
Photoelectric ceUs: 707
Photoelectric colorimeters: .see Colorimeters
and Spectrophotometers
Photoemissive cells: 700, 707 •

Photometer: 693
method, 700, 707
Photomultiplier tube: 707
Pipettes:
automatic, 78
calibration of, 77
Dafert, 103
draining time of, 77
filler, 77

graduated, 76, 78
Lunge- Rey, 81, 307
manipulation of, 76
safety, 77
tilting, 78
tolerances of, 76
transfer, 76
Planimeter; 214
Plastic apparatus: 85
table of plastics, 85
Platinum apparatus: 86, 527
care and use of, 86
cleaning and preservation of, 86
Platinum-clad stainless steel ware: 88
Platinum, D. of: as element, (g) 474
Poggendorff ’s compensation method: 579
Polarisable micro-electrodes: 635
Polarisation e.m.f.: 517
Polarised indicator electrodes: 516
Poiarograras: 633, 639
derivative, 648

of air-saturated potassium chloride solution,
656

Polarograpb: 633
manual non-recording, 647
commercial instruments, 648, 662
Polarographic analysis: ancillary equipment for,
650

applications of, 650

basic apparatus for, 633, 634, 657

cells for, 652

controlled current methods, 661
D. of cadmium in solution, 653
D. of lead and copper in steel, 656
D. of formula and stability constant of a
metal ion complex by, 641

917

INDEX

Polarographic analysis—

D. of two or more ions by, 644
evaluation of quantitative results in, 644, 646
general introduction to, 632
influence of dissolved oxygen in, 643, 655
investigation of, 655
quantitative technique of, 642, 663
supporting electrolyte, 636
theoretical principles of, 636
Polarographic maxima: 638
due to oxygen, 643
suppression of, 638, 653
Polarographic wave: 635, 891
equation of, 637, 641
half wave potentials, 638, 653
table of, 891

Polarography: alternating current, 633, 656
ancillary equipment for, 650
capillaries for, 651
cells for, 652
comparative, 661
direct current, 633, 636
dropping mercury electrode, assembly, 650
Ilkovic equation, 637
linear sweep, 650
oscillographic, 656, 660
pulse, 658

Randles-Sevcik equation, 661
rapid, 650
square wave, 658
wave heights, 645
‘Policeman’ •■94
Polycarbonate apparatus; 85
Polymethylpentene (TPX) apparatus: 85
Polypropylene apparatus: 85
Polystyrene apparatus: 85
Polytetrafluoethylcne (teflon) apparatus: 85
Polythene apparatus: 83, 85
Porcelain apparatus: 84
filtering crucibles, 96
Post-precipitation : 4 1 0
Potassium, D. of: as chloroplatinate, (g) 476
as dipotassium sodium

hexanitrotocobaltate(lll)

(cobaltinitrite), (g) 475
by flame photometry, 837
in admixture with sodium, (fl) 837
with tetraphenylboron, (am) 681, (g) 475, (ti)
347

Potassium bromate solution: analyses involving,
390

indicators for, 391
oxidising properties of, 391
preparation of 0. 1 A^, 392
Potassium chloride (m'trate) - agar bridge: 605

Potassium cyanide: standard solution of, 61 1
use in D. of cobalt, (p) 61 1
use in D. of nickel, (p) 61 1
Potassium cyanoferratc(ll): D. of, (ti) 370
Potassium cyanoferrate(III): D. of, (ti) 385
Potassium cyanonickelate(II):prepn., 323
Potassium dichromate solution: analyses
involving, 359
oxidising properties of, 359
internal indicators for, 359
preparation ofO.ltV, 360
redox indicators for, 359
standardisation of, by iron, (cm) 559, (ti) 360
standardisation, by iton(II) ethylene
diammonium sulphate, 360
Potassium ferricyanide; see Potassium
cyanoferrate(IIl)

Potassium fertocyanide: see Potassium
cyanoferrate(ir)

Potassium hydrogeniodate : 306
Potassium hydrogen phthalate: as pH standard,
586,886

as standard substance, 236 . .

Potassium iodate solution: analyses involving,
386

detection of end points in titrations, 387
for standardisation of acids, 385
oxidising properties of, 386
preparation of 0.025 Af, 387
Potassium iodate and potassium iodide: 372
standard solution of, 372
standardisation of strong acids with, 372
Potassium iodide: standardisation of potassium
permanganate solution by, (p) 610
Potassium nitrate: D. of, (s) 759
Potassium nickelocyanide: preparation of, see
Potassium cyanonickelate(ll)

Potassium permanganate solution: analyses
involving, 348

applications in alkaline solution, 349
discussion on standardisation of, 350
oxidising properties of, 348
permanence of, 350
preparation of 0. 1 Al, 35 1
standardisation by arsenic(III) oxide, 351 '
standardisation by ethylenediammonium
iron(II) sulphate, 353
standardisation by iron, (cm) 560, (ti) 353
standardisation by potassium iodide, (p) 610
standardisation by sodium oxalate, 352
Potassium tetracyanonickelate: 267, 323
Potassium thiocyanate solution: prepn. of 0.1/7,
341

standardisation of, 341
use of, 160, 342,741

918

INDEX

Polential mediators; 295

Potentials: calculation of standard (reduction),

52 • ■ ' • ■

decomposition', 517 -

deposition, 523

electrode, 46, 566 ■ • '

formal, 291

half-wave, 638 • ^

liquid junction, 567 '

table of, 47

stmdard(or reduction), 51 •'

Potentiometer: 579 ' ’ ,

commercial, 581,594

Potentiomettic titrations: 5, 566, 567, 591
automatic, 594
classical method, 571

derivative method for end points, 596, 598
differential method, 573
electronic instruments, 580
general considerations, 591
theory of, 588, 591

with mercury electrode and EDTA, 607
Potentiometric titrations, experimental details
for: 1

complexation reactions, 601
neutralisation reactions, 600, 602
oxidation-reduction reactions, 599, 601, 603
precipitation reactions, 601, 603
Potendometry: 566
direct, 566, 585
null point, 567
Potentiostats: 649
Precipitants: organic, 419
Precipitate: ageing of, 107
digestion of, 410
drying and ignition of, 112
effect ofacids upon, 32
formation of, 408
ignition of, 415

incineration of filter paper alone, 114
purity of, 409

solvent and solubility of, 32
temperature and soly. of, 33
washing of, 112,413
Precipiution: 106
andsupeisaturation ,408
completeness of, 404
conditions of, 410
co-precipitation, 409

effecu^of common ion upon completeness of,
successful analysis by,

fractional, 29, 415

from homogeneous solution, 41 1

Precipitation — continued

aluminium as basic succinate, 449
aluminium as oxinate, 436
barium as chromate, 460
barium as sulphate, 453
calcium as oxalate, 354, 458
chromium as lead chromate, 460
iron as formate,' 443
of hydroxides, 417
of sulphides, 416
post-,410

practical points concerning, 41 1
Precipitation methods: 403
Precipitation reactions: 224, 226, 336
theory of, 279, 601
Precision: 10, 120
Preparation for analysis: 103
Preventive solution: 348
Primary standard substances: requirements of,
235

Prisms, for spectrophotometers: 71 1

Probability curve: see Normal distribution

Propylarsom'c acid : 427

Protective colloid : 407

Purification of substances: 99

Pyridine : 427, 457, 473, 489

Pyrogallol:428, 451,456

Pyrolusite: D. of manganese content of, 607

Quenching: 774
Quinaldic acid: 428, 457, 488
Quinaldine Red indicator: 242
Quinine, D. of: (fu) 777
Quinoline molybdopbosphate: 314

Radioactivity; 6
‘Raies ultimes’ : 788
table of, 892

Randles Sevcifc equation: 661
Reagents: 98
analytical, 98

approx. pH of some common solutions of,
887

quality of, 98

saturated solns. of at 20 °C, 882
spectrographically standardised, 885
sources of analysed standards, 885
use of, 98

see also Electrolytically generated reagents
Recording of results: 58, 307
Rccrystallisation of solids : 99
Redox indicators: preparation and properties
of, 293
table of, 294

Red rod immersion heater; 89

919

INDEX

Redox processes: 53, 228, 288, 348
Reducing agents : 228
equivalent of, 231
table of, 233

Reductant: equivalent weights of, 33
Reduction: 231
by chromiuin(II) salts,
by hydrogen sulphide, 401
by Jones reductor (zinc amalgam), 395
by liquid amalgams, 397
by silver reductor, 388
by sulphurous acid, 400
by tin(II) chloride, 399
see also Iron(III), reduction of
Reduction potentials: 5 1
Reference electrodes: potentials of common,
890

Relative error: 127
mean deviation, 127
Releasing agents: 822
Residual current: 636
Resistance: 515
Resistivity: 615
Resonance line sources: 822
Results: comparison of, 130
Reverse osmosis: 82
Revalues: 194
determination of, 194, 199
Rider: 60
Rose crucible: 461
Rotated electrode: 527, 530
R.U. powder: 788, 796

Safety: during sampling, 138
in atomic absorption spectrophotometry, 833
in gas chromatography, 218
in the laboratory, 58
pipette, 77

Sah'cylaldebyde oxime: 424, 468, 473
Salt bridge: 546, 604, 605
Salt effect: 26

Salts: pH of hydrolysed solns. of, 39
Samples: crushing and grinding of, 137
dissolution of, 104, 832
weighing of, 64, 69
see also Analysed samples
Sampling: 135
hazards of, 138
of gases, 135, 210
of liquids, 136, 210
of solids, 136

Saturated solutions of some reagents: (table) 882
S.C.E. (Saturated calomel electrode): 49, 527,
569, 890

Schoniger oxygen flask: see Oxygen flask
Schwarzenbacb classiScation : 258, 26 1
Screened indicators: 243
Screens: perforated for crucibles, 1 15
Selective precipitation : 8
Selectivity : of analytical methods, 8
in EDTA titrations, 267
Secondary pH standards: 886
Selenium, D. of: as element, (g) 477
Sensitivity: (fl) 834
Separation coefficient: 144, 173
Separations; by chromatographic methods, 10,
193, 202

by coulometric methods, 548
by EDTA methods, 267
by electrogravimetric methods, 522, 524
by ion exchangers, 172
by polarography, 656
by precipitation methods, 41 5
by solvent extraction, 144, 151
gravimetric, 419
Shaker; 95

SI um'ts: xxvii, 404, 615, 695
Significant figures: 125
SUica apparatus; 84

Silica, D. of: as quinoline molybdosilicate, (g)
503

by molybdenum blue method, (s) 757
in an ‘insoluble’ silicate, (g) 502
in a ‘soluble’ silicate, (g) 501
Silica filtering crucibles : 85, 1 1 3
Silicate:

analysis of ‘insoluble’, 502
analysis of ‘soluble’, 501
Silver apparatus: 88
Silver, D. of : as chloride, (g) 479
as metal, (eg) 537

by ammonium thiocyanate, (ti) 342
by EDTA, (ti) 323

by 1,10-phenanthrolineand bromopyrogallol
red, (se) 161
in alloys, (ti) 342
by turbidity method, 286
Silver nitrate solution: prepn. ofO.EK, 336
standardisation of with sodium chloride and
potassium chromate indicator, 336
standardisation of with sodium chloride and
adsorption indicator, 338
Silver reductor: 398
prepn. of silver for, 398
use of, 367, 399

Silver-silver chloride electrode: 49, 571, 890
Simple gravimetric determinations: 432
Sintered-glass filtering crucibles: 96
advantages of, 1 10

920

INDEX

Soda lime: 492

Sodium, D. of: as sulphate, (g) 479
as line uranyl acetate, (g) 480, (ti) 324
by flame photometry, 837
indirectly by EDTA, 324
Sodium arsenitesoIn.:prepn; of standard, 352
ieen/ioArsenicCIII) oxide
Sodium bismnthate: 358 ■ ,

Sodium carbonate: as standard substance, 298
choice of indicators for, 299
content of washing soda, (ti) 306 -
preparation of pure, 298
titration with strong acids, (hf ) 630
Sodium diethyldithiocarbamate: 15i, 155,734
Sodium diphenylaminesuiphonate: 359, 360, 363
Sodium hexanitrotocobaltate(III)

(cobaltinitrite): 475

Sodium hydroxide solution: prepn. of carbonate
free. 302

prepn. by ion exchange method, 302
prepn. of 0.1 3/, 304
pH of 0.1 Af, 887- .
standardisation of

with standard acid, (hf ) 630, (ti) 304
with potassium hydrogenphthalate, (ti) 305
Sodium molybdate: 315
Sodium oxalate: 366, 630
Sodium pentacyanoferrate: prepn. of, 199
Sodium peroxide: analysis of, (ti) 355
Sodium starch glycollate: 373
prepn. and use of the indicator soln., 374
Sodium tetraborate: as standard substance, 300
Sodium tetraphenyl boron : 429, 68 1
reagent, 429, 475
recovery of, 476

Sodium thiosulphate solution; prepn. ofO.lA
374

stability of, 374

standardisation of
bycerium(IV) sulphate, (ti) 377
by potassium dichromate, (ti) 376
by potassium iodate, (ti) 375
by potassium permanganate, (ti) 376
by standard iodine soln., (am) 683, 685,
(ti) 377
uses of, 379
Sodium tungstate: 753
Sodium zinc uranyl acetate : 323

of: by EDTA, (ti) 333
ack:272,745

bolochrome Black 68:273

olochrome Cyanine R: 729
^jo^ome Dark Blue: 273
0 ‘bbes of inorganic substances: table of, 883

Solder analysis

Solochromp Bi.

Solubility product : 25
calculations involving, 26
importance of, 27
principle limitations of, 26
Solution of sample: 104
experimental details for, 104
Solvent extraction, liquid-liquid systems: 143
backwashing in, 153
batch, 151

choice of solvent for, 151
completeness of, 1 53
continuous, 153
distribution coefficient in, 143
factors favouring, 145
general discussion of, 9
quantitative treatment of equilibria in, 146
reagents for, 149
separation factor in, 144
simple applications of, 269
some practical considerations, 151
special separatory funnel for, 152
species in, 145
stripping in, 153
Sorbitol: 311

Sources of analysed samples: 885
Spark source: 793
Specific absorption coefficient: 697
Specific extinction coefficient: 697
Specific gravities, tables of: acids at 20 °C, 880
alkaline solns. at 20 “0,881
Spectrograph: adjustment of, 801

commercial instruments, 789, 791, 800
Spectrographic analysis: see Emission
specirographic analysis
Spectrographically standardised substances :

addresses of suppliers, 885
Spectrophotometer: commercial instruments,
721

double-beam, 725
method, 703
operation of, 721, 725
single-beam, 720
see also Colorimeters

Spectrophotometric determinations: 683, 727
D. of absorption curve, 759
D. of conen. of potassium nitrate, 759
D. of simultaneous, chromium and
manganese, 763
pK value of an indicator, 76 1
Spectrophotometric titrations: 767
apparatus for, 768
examples of, 769, 770, 771
Spectrophotometry: theory of, 6, 695
Spectrum: 694

examples of spectrographic emission, 793, 797

921

INDEX

Sficttam— continued
projector, 794
visible, 694
Stability constants:

of complexes, 258, 263, 264, 641
table of, 264
Stainless steel ware: 88
Standard addition: method of, 125, 645, 655,

831

Standard curves: in spectrophotometry, 728
Standard deviation: 127
Standard potentials: 47, 51, 52
Standard series method: 699, 701
Standard solutions: 100, 223, 225, 3 1 7, 336
for pH: 587, 694
prepn. of, 100,235, 296, 832
storage of: 101

Standard substances for acidimetry and

alkalimetry: anhydrous sodium carbonate,
236

barium hydroxide, 306
benzoic acid, 236, 305
borax, see Sodium tetraborate
constant b.p. hydrochloric acid, 236, 297
potassium hydrogeniodate, 236, 306
potassium faydrogenphthalate, 236
sodium hydroxide, 301, 304
sodium tetraborate, 236
succinic acid, 305
sulphamicacid, 306

Standard substances for prcdpilation titrations:
potassium chloride, 236
silver, 236
silver nitrate, 236
sodium chloride, 236
Standard substances for redox titrations:
arsenic(III) oxide, 350, 366, 378 ■
iodine, 236
iron, 236

iron(ll) ethylenediammoniura sulphate, 351

potassium faromate, 236

potassium dichromate, 236

potassium hydrogeniodate, 236

potassium iodate, 236

potassium iodide, 610

sodium oxalate, 236

Standard temperature: for graduated glassware,
72

Starch indicator solution: 373
disadvantages of, 373
prepn. and use of, 374
Starch-urea indicator: 374
Stationary phase: 10, 210
Statistical methods in analysis: 127
Steam baths: 89

Stirring: combined with heating, 95
during electrolysis, 530
magnetic, 94
of liquids, 94
rods, 94

variable speed, 95
Stoichiometric end point; 236
Storage of solutions for titrimetric analysis: 100'
Strontium, D. of: as hydrogen phosphate, (g)
481

as sulphate, (g)481

Strengths of common acids and ammonia : 88 1
Student’s t-test: 130
tables for, 895

Sublimation: purification of solids by, 100
Substitution titrations: 320
Succim'c acid: 305, 449
Sulpbamic acid : 306
in electro-analysis, 532
in homogeneous precipitation, 413
Sulphanilimide: 757

Sulphate, D. of: as barium sulphate, (g) 409,

504

(i) filter crucible method, 507

(ii) filter paper method, 506 ■

(iii) with agar-agar as coagulant, 505
by EDTA, (ti) 335

by nephelometry, 786
indirectly, (aa) 843
with barium chloranilite, (s) 704
with lead nitrate, (am) 677
Sulphides, D. of: as barium sulphate, (g) 507
by iodine, (ti) 384
by potassium iodate, (ti) 385 '
in minerals, (g) 507

Sulphites, D, of: as barium sulphate, (g) 510
by iodine, (ti) 383
Sulphonephthaleins; 242
Sulphur, D. of: in iron pyrites, (ti) 187
in mineral sulphides, (g) 507
by dry process, 508
by wet process, 509

with the aid of an ion exchanger, (ti) 1 87 •

Sulphuric acid, D. of: in the concentrated acid,
(ti)307

in oleum, (ti) 307

Sulphurous acid and sulphites, D. of: as barium
sulphate, (g) 510
by iodine, (ti) 383
reductions with, 400

SupersaturatioD and precipitate formation: 408
Supporting electrolyte; 547, 636, 683
Supports: for partition chromatography, 210
Suspensoids: 405

Systematic gravimetric analysis : 449

922

INDEX

t-iest:I30 • ' - •

tables for, 895
Tannic acid: 426
Tantaluinboat:819

Tartrazine indicator: 341 ,

‘TeepoT detergent: 73
Teflon apparatus: 85
TelluiitK, D. of: (ti) 368

TelluriunijD.of: , • ■

as element, (g) 477

by ceiium(lV) sulphate, (ti) 368 • ■ . ,

in admixture with selenium, (g) 477
Temperature: corrections for graduated flasks,
73 - •

standard for graduated glassware, 72
Temperature programming : for gas , :

chromatography, 214
Test papers: 317

Tetraphenylarsoniiim chloride: 429, 482
Thallium, D. of: as tetraphenylarsonium
chiorothallate, (g) 482
as thallium(I) chromate, (g) 482
by potassium iodate, (ti) 389
Thenoyltrifluoroacetone: 150 .

Thermal analysis: 849
Thermal conductivity detector: 21 1

Thermobalance: 849, 853

Thermogravimetric curves : 849
Thermogravimetry (TG): 849
applications, 855, 858
experimental
calcium oxalate, 858
copper sulphate, 858
experimental factors, 851
instrumentation, 853
introduction, 849,

Therraometric titrations: 864
applications, 866

experimental details, chloride ion, 867 •
instrumentation, 866

Thin layer chromatography: 193, 196 , 200
see also Chromatography
Thiocyanates, D. of: as barium suiphate, (g) 5 10
^copper(l) thiocyanate, (g) 510
oysilvernitrate,(ti)339,

acid): 743,

Thionalide: 471

TIuosulphate, D. of: as barium sulphate, (g)
assilversulphide,(g)51I

J ‘odine, (am) 683, 685, (cm) 558, (ti) 37
J afro Sodium thiosulphate

oxalate, (g) 483, (hf) 630

Thorium, D. of — continued
via sebacate, 483
Thorium chioraniliate: 755
Thymol Blue: 240
Thymolphthalein: 242, 244
complexone (thymolphthalexone), 277
Tin, D. of: as dioxide, (g) 484
by N-benzoyl-N-phenylhydroxylamine, (g)
484

by cupferron, (g) 484
by dithiol, (s) 748
by potassium iodate, (ti) 389
in bearing metal, (eg) 539
in canned foods, 749
in solder, 333
with lead, 333

Titanium, D. of; as oxide, via 4-hydroxy-
phenylarsonic acid complex, (g) 485
via tannic acid and phenazone complexes,
(g)485

by hydrogen peroxide, (s) 750
in brass, (em) 803
_j]Citan Yellow: 745
Titrand:223
Titrant: 223
"Titration: 223

in an inert atmosphere, 353, 361
Titration curves: of acid-base reactions, 244
of amperometric titrations, 674
of complex formation reactions, 265
of conductometric titrations, 622
of EDTA reactions (pM curves), 265
of high-frequency titrations, 630
of neutralisation reactions, 248, 249, 252
of oxidation-reduction titrations, 289, 290
of precipitation reactions, 28 1
of spectrophotometric reactions, 767
Titration error: 223
Titrations: acid-base, 623
amperometric, 672
atomic absorption inhibition, 843 .
automatic, 594
biamperoraetric, 685
colorimetric, 767
coraplexation, 257, 624
conductometric, 621
coulometric, 556
dead-stop end point, 685
displacement, 253, 623
EDTA, 266,317, 324
high-frequency, 626
oxidation-reduction, 348, 625
potentiometric, 59,1
precipitation, 279

923

INDEX

Titrations— co/)//««erf
recording of, 307, 594
spectrophotometric, 767, 769, 771
see also under individual liiralion
Titrimetric anaiysis: 5, 223
calculations of, 234, 300, 307
classifications of reactions in, 224
conditions a reaction must fulfil for^ 223
general discussion of, 223
storage and preservation of solns. for, 101
technique of, 299

Titrimetric apparatus: see Graduated glassware
Toluene-3, 4-dithioI : see Dithiol
Tongs: for crucibles and beakers, 9 1
Transmittance: 696
Triaogulation: 215
Triethanolamine: 273
Tri-n-butyl phosphate: 151
Triethyl phosphate: in homogeneous
precipitation, 412
Triethylcnetetranune-A^,At,iV'

,yV",A''",A^'"-hexa acetic acid: (TTHA), 262
Trimethyl phosphate: in homogeneous
precipitation, 412
Tri-n-octyl phosphine oxide: 151
Triton X-100: 638
Tropaeolin 0: 240, 243
Tropaeolin 00: 240, 243
Tungsten, D. of: as barium tungstate, (g) 486
as trioxide, via tannic acid and pbenazone,
(g)486

bydithiol,(s)751
in steel, 751

Turhidimetric analysis: determinations by, 784
general discussion on, 781
instruments for, 781
Turbidity method : for D. of silver, 286
TyndaUcflrect:404,781

Ultramicroscope: 404
Ultraviolet radiation: 694
Ultraviolet/visible spectropbotoineters: see
Spectrophotometers and
Spectrophotometry
Units: of length, 404, 695
of mass, 61
of volume, 71

Universal buffer solution : 886
‘Universal’ indicators: 244
Uranium, D. of: by cupferron, (g) 487
by ceriumllV) sulphate, (li) 370
by liquid anion exchanger, 189
by oxine, (g) 487, (se) 163, (ti) 393
sepn. of, plus cobalt, (ch) 188

Uranyl zinc acetate: 480

Urea: in homogeneous precipitation, 412

Vanadate, D. of: (ti) 390
Vanadium, D. of: as silver vanadate, (g) 489
by hydrogen peroxide, (s) 752
by potassium iodate, (ti) 390
by phosphotungstate method, (s) 753
in lubricating oil, (aa) 840 •
in steel, 755

Variamine Blue B: 279, 322
Variance: 128

Variance ratio test: see F-test
Vibro-spatula: 104

Vinegar: D. of acetic acid content of, (ti) 307
Volatilisation method: 430
Volhard procedure: applications of, 342
experimental details of, 342
theory of, 340
Volt: 515

Voltaic cells : 49, 5 1 6 ’

Voltammetry: 5, 632
anodic stripping, 664
voltammogram, 664
Volume distribution coefficient: 174
Volumetric analysis: see Titrimetric analysis
Volumetric apparatus: see Graduated glassware
Volumetry: 5
Vycor apparatus: 85

Walpole technique for colour matching: 703
Wash bottles: 83
polythene, 83

Wash solutions; for precipitates, 415
Washing of the precipitate: 1 1 1, 413
by decantation, 112
solubility losses in, 1 12, 414
Washing soda: D of sodium carbonate in, 308
Water: absorbents for, 492
ammonia-free, 730
deionised, 82
D. of hardness, (h) 328
D. of total cation concentration, 183
D. with Karl Fischer reagent, 688
distilled, 82
ionic product of, 33
table of volume of 1 g at various
temperatures, 75
typesofandpH, 82
weight to give 1 dm* at 20 °C, 75
Water baths: 89

Water of hydration : D. of, (g) 432
Karl Fischer method, 689

924

INDEX

Wjieheight-concentration plots: 644
Waiebeighls: measurement of, 645
Wafelengtiis: approximate of colours, 694
limits of various types of radiation, 694
ranges of various light filters, 7 1 0
sensitivity of eye to various, 709
units for, 695
Wjienii!nbcrs:695
H'ajie-lfefrWdge; 620
Weil electrolytes: 17
Weighing: bottles, 97
effects of buoyancy of air upon, 70
errors in, 69

externally controlled weight loading, 63
chemical samples, 104
reduction of to vacuo, 7 1
funnel, 97

with riderless aperiodic balance, 62
with substitution balances, 64
with two-knife single pan balances, 63

Weights: 61
calibration of, 62
standard, 61
tolerances (NBS), 62

tolerances and accuracy of certification
(NPL),62

Weight of water to give one dm^ : 75

Wheatstone bridge: 617, 620, 625
Whitemetal: analysis of, 807

Wiping, electrification due to: 69
Wittplate: no

Wood's metal: analysis of, (ti) 333

X-ray fluorescence: 6
X-rays: 694
Xylene Cyanol FF : 243
Xylenol Orange: 276
Xylidine Ponceau : 39 1

Zimmermann-Reinhardt process: 348
solution, 348
use of, 348

Zinc, D. of: as 8-hydroxyquinaldate, (g) 489
as pyridine thiocyanate, (g) 489
as quinaldate, (g) 488
by EDTA, (am) 680, (p) 6 10, (ti) 330, 393
by oxine, (fu) 779, (ti) 393
by oxygen flask, 1 16
by fluorimetry, (fu) 779
in brass, (em) 803

with manganese and magnesium, (ti) 329
Zinc sepn. of: plus cobalt, copper, nickel, (ch)
199

plus cobalt, manganese, nickel, (ch) 200
Zinc amalgam: prep, of, 397
see also Liquid amalgams
Zinc and magnesium: sepn. of by ion exchange,
181

Zinc uranyi acetate: 480
Zincon:281

Zirconium, D. of: as dioxide, (g) via basic
selenite, 490

as dioxide via mandelate, 490
Zirconium, extraction of: 178
Zone refining: 100
Zwitterions: 262

925

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Texts

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Vogels Textbook Of Quantitative Inorganic Analysis"

See other formats